Citation
Parameters governing airway pressure homogeneity during low frequency jet ventilation

Material Information

Title:
Parameters governing airway pressure homogeneity during low frequency jet ventilation
Creator:
Partile, Joshua Daniel
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Bioengineering, CU Denver
Degree Disciplines:
Bioengineering
Committee Chair:
Smith, Bradford
Committee Members:
Fink, Daniel
Hunter, Kendall S.

Notes

Abstract:
Low frequency jet ventilation (LFJV) is a widely used form of ventilation that currently has no defined safe pressures or flow rates because the methodology for safe and effective ventilation has been insufficiently researched. This study looked at pressure in the distal airways during low frequency jet ventilation and how different parameters contributed to an even pressure distribution across the lung (pressure homogeneity). Two physical models of airway trees were fabricated: one based on a CT scan of a healthy 17-year-old male and second that was a symmetric, idealized representation of the human airway tree. We found that needle placement closer to the proximal end of the trachea and centered within the airway cross section produced the most homogenous pressures at the 5th airway generation. Additionally, the more distal the needle placement was in the trachea, the more sensitive the distal airway pressure was to the needle being off the tracheal centerline. The distal airway pressure is determined by total entrained flow rate and not just pressure supplied to the ventilation needle. The relationship between flow rate and distal airway pressure is a second order polynomial with R2 > .99. Supply pressure at the needle is a determining factor in flow rate but in itself is not sufficient to determine distal airway pressure. A device was created to measure the total entrained flow rate and the design considerations are detailed herein. The distal airway pressures during ventilation with a clinical laryngoscope and subglottiscope were also measured. The depth of scope placement from one to three centimeters did not substantially affect distal airway pressure or homogeneity. The placement of the needle in the laryngoscope substantively changed distal airway pressure. Even with consistent needle depth, a needle angle change of a few degrees nearly doubled the distal airway pressures. We conclude that laryngoscope designs should be updated to securely hold the ventilation needle to provide consistent airway pressures. Furthermore, the jet flow rate should be used to set inlet pressure prior to ventilation so that distal airway pressures are consistent across different ventilation hardware configurations.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Joshua Daniel Pertile. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
PARAMETERS GOVERNING AIRWAY PRESSURE HOMOGENEITY DURING LOW
FREQUENCY JET VENTILAITON By JOSHUA DANIEL PERTILE
B.S., University of Wisconsin, 2017
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science in Bioengineering Bioengineering Program
2019


© 2019
JOSHUA DANIEL PERTILE
ALL RIGHTS RESERVED


This thesis for the Master of Science Bioengineering degree by Joshua Daniel Pertile has been approved for the Bioengineering Program by
Bradford Smith, Chair Daniel Fink Kendall Hunter
Date: July 27th, 2019


IV
Pertile, Joshua Daniel (M.S., Bioengineering)
Parameters Governing Airway Pressure Homogeneity During Low Frequency Jet Ventilation Thesis directed by Assistant Professor Bradford Smith
ABSTRACT
Low frequency jet ventilation (LFJV) is a widely used form of ventilation that currently has no defined safe pressures or flow rates because the methodology for safe and effective ventilation has been insufficiently researched. This study looked at pressure in the distal airways during low frequency jet ventilation and how different parameters contributed to an even pressure distribution across the lung (pressure homogeneity). Two physical models of airway trees were fabricated: one based on a CT scan of a healthy 17-year-old male and second that was a symmetric, idealized representation of the human airway tree. We found that needle placement closer to the proximal end of the trachea and centered within the airway cross section produced the most homogenous pressures at the 5th airway generation. Additionally, the more distal the needle placement was in the trachea, the more sensitive the distal airway pressure was to the needle being off the tracheal centerline.
The distal airway pressure is determined by total entrained flow rate and not just pressure supplied to the ventilation needle. The relationship between flow rate and distal airway pressure is a second order polynomial with R2 > .99. Supply pressure at the needle is a determining factor in flow rate but in itself is not sufficient to determine distal airway pressure. A device was created to measure the total entrained flow rate and the design considerations
are detailed herein.


V
The distal airway pressures during ventilation with a clinical laryngoscope and subglottiscope were also measured. The depth of scope placement from one to three centimeters did not substantially affect distal airway pressure or homogeneity. The placement of the needle in the laryngoscope substantively changed distal airway pressure. Even with consistent needle depth, a needle angle change of a few degrees nearly doubled the distal airway pressures. We conclude that laryngoscope designs should be updated to securely hold the ventilation needle to provide consistent airway pressures. Furthermore, the jet flow rate should be used to set inlet pressure prior to ventilation so that distal airway pressures are consistent across different ventilation hardware configurations.
The form and content of this abstract are approved. I recommend its publication.
Approved: Bradford Smith


To all those who instilled in me a sense of curiosity, especially my parents


VII
Acknowledgements
I would like to acknowledge Bradford Smith for scientific and material support and mentoring me though the thesis process, Daniel Fink for funding and advice, Jennifer Wagner and Emily DeBoer for providing the CT airway model and assisting in the fabrication of the idealized airway model, and Michelle Mellenthin for countless hours of help with programming and circuit design and supplying hundreds of cups of high quality coffee.


VIII
Table of Contents
CHAPTER
I. INTRODUCTION..............................................................1
II. BACKGROUND................................................................7
III. MATERIALS AND METHODS...................................................12
Model Airway Pressure Measurements......................................12
Jet Needle Positioning System..........................................15
Signal and Data Processing.............................................17
Airway Pressure Measurement Procedure..................................21
Flow Measurement........................................................22
Additional Considerations..............................................26
IV. RESULTS..................................................................29
Flowmeter Characterization..............................................29
Airway Pressure Measurements............................................30
Data Normalization......................................................33
Airway Pressure Relationships...........................................35
Airway Pressure Distribution............................................40
Laryngoscopes and Distal Airway Pressure
53


IX
V. DISCUSSION.............................................57
The Relationship Between Inlet Pressure. Flow Rate, and Mean Distal Airway Pressure ..................................................................................58
Jet Flowmeter Development and Characterization.................................61
Jet Needle Position and Distal Airway Pressure Heterogeneity...................62
Effects of a Laryngoscope and Subglottiscope on Distal Airway Pressure.........63
Conclusions and Suggestions for Clinical Practice..............................65
Future Investigations..........................................................66
REFERENCES..............................................................................67
72
APPENDIX


X
List of Figures
Figure 1: The anatomy of the human lung [5]......................................2
Figure 2: Air entrainment in a laryngoscope......................................5
Figure 3: A standard jet ventilation setup. The red arrow is pointing to the pressure regulator. 8
Figure 4 : Idealized airway geometry based on the morphometry models from Florens,
Horsfield, and Weibel [46-48]....................................................13
Figure 5: Model of the human airway tree, derived from the CT scan of a 17-year-old healthy
human subject....................................................................14
Figure 6: Photo of the experimental setup with the CT airway block. The regulator is indicated
with a red arrow....................................................................16
Figure 7: Wiring diagram for the data acquisition system from the pressure transducers to the
computer............................................................................19
Figure 8: A top-down view of the trachea with the thirteen ventilation positions shown: The
center, and three steps in each direction...........................................22
Figure 9: Venturi flow meter design. The locations of Ai, kj, Pi, and P2 (Eq. 1) are labeled in the
cutaway (A)..............................................................23
Figure 10: Jet needle mounting system..........................................24
Figure 11: The curve of the flow rate vs the measured voltage. This provides an accurate
measurement from a flow rate of 0.25 L/s to approximately 3.25 L/s.........26
Figure 12: Twist on Mounting Bracket..............................................27
Figure 13: Functioning prototype of the air flow entrainment measuring device, with housing


XI
removed.........................................................................28
Figure 14: Volume determined as the integral of the flow measured using the venturi flowmeter (blue circles). The target value for each measurement is 3 L (red line) that is delivered
using a calibrated syringe......................................................30
Figure 15: The demultiplexed, unfiltered, uncalibrated voltage readings from each sensor during a ventilation test run on the CT airway block. The needle was centered and moved
through all 7 depths. D1-D7 denote the depth into the trachea from 1cm to 7cm...31
Figure 16: Distal airway pressure measurements in the CT airway block with the jet needle 6mm from the left wall of the trachea (first step to the left) showing the filtered pressure for all 16 distal airway sensors (a). The homogeneity graph (b) shows the mean pressure (heavy black line) and the shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. Note that (b) is normalized by the
inlet pressure. D1-D7 indicate depth into the trachea from 1cm to 7cm...........33
Figure 17: Transient in inlet pressure over the three second ventilation burst provided during the experiment. This is measured by a pressure transducer placed immediately before
the ventilation needle..........................................................34
Figure 18: Relationship between distal airway pressure and total flow rate including entrained air. This is in the center position of the idealized airway block at 1 (a), 4 (b), and 7 cm (c)
into the trachea................................................................36
Figure 19: Relationship between distal airway pressure and total flow rate shown at 1, 4, and 7
cm into the trachea, with the needle centered in the idealized airway block.....37
Figure 20: Distal airway pressure vs inlet pressure for the center needle placement in the


xii
idealized airway model at depths of 1 (a), 4 (b), and 7 cm (c)..................38
Figure 21: The linear regression of inlet pressure vs distal airway pressure for the CT airway
model with the needle centered in the trachea at depth 1........................39
Figure 22: The linear regression of inlet vs distal pressure for the idealized airway model with
the needle centered in the trachea at depth 1...................................40
Figure 23: The numbered positions of the sensors on the idealized airway block.........41
Figure 24: Sensor represented on the homogeneity grid..................................42
Figure 25: Example of airway bifurcations to reach sensor 9. Each darker shade represents a
branch farther into the block...................................................43
Figure 26: Comparison of homogeneity in the distal airways for the idealized airway model at
depths 1 (a), 4 (b), and 7 cm (c)...............................................44
Figure 27: The homogeneity grids at depths 1 (a), 4 (b), and 7 cm (c) with the jet needle
centered in the trachea in the CT airway block..................................45
Figure 28: Trimodal ventilation demonstrated in the CT airway block with the needle placed in
the far right position in the trachea at depths of 1 (a), 4 (b), and 7 cm (c)...47
Figure 29: The homogeneity grid at depth 1 with an 18 gauge needle in the center of the
trachea for the CT airway tree..................................................47
Figure 30: Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left side of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range
from the highest to lowest mean value for an individual sensor at that depth. The steps


XIII
toward the wall are numbered from the center of the trachea out. Left 1 is closest to
the center and left 3 is closest to the tracheal wall.............................49
Figure 31: Comparison of needle placements in the idealized airway block for the center and
the placements closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the
highest to lowest mean value for an individual sensor at that depth...............50
Figure 32: Homogeneity in the distal airways of the CT airway block as a function of distance
from the left side of the trachea. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to
the tracheal wall.................................................................51
Figure 33: Comparison of needle placements in the CT airway block for the center and the
placements closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the
highest to lowest mean value for an individual sensor at that depth...............52
Figure 34: Pressure homogeneity grids for the subglottiscope with the orifice 1 and 3
centimeters past the laryngeal folds. This is in the CT airway block..............54
Figure 35: The homogeneity grids for the Ossoff-Pilling Laryngoscope ventilation. The needle is


XIV
mounted at the bottom of the right wall and tilted up at angles of 10, 5, 3, and 0
degrees. This is the CT airway model.......................................55
Figure 36: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the dorsal wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Dorsal 1 is closest to
the center and dorsal 3 is closest to the tracheal wall....................72
Figure 37: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the ventral wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Ventral 1 is closest
to the center and ventral 3 is closest to the tracheal wall................73
Figure 38: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center


XV
and left 3 is closest to the tracheal wall....................................74
Figure 39: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the right wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Right 1 is closest to the
center and right 3 is closest to the tracheal wall............................75
Figure 40: Homogeneity in distal airway pressure in the CT airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the
highest to lowest mean value for an individual sensor at that depth...........76
Figure 41: Homogeneity in distal airway pressure in the idealized airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range
from the highest to lowest mean value for an individual sensor at that depth..78
Figure 42: Homogeneity grid for ventilation in the idealized airway block with the needle placed
in the far right position in the trachea at depth 1...........................79
Figure 43: Homogeneity grid for ventilation in the idealized airway block with the needle
placed in the far right position in the trachea at depth 4. Bimodal ventilation is seen


XVI
here..................................................................................80
Figure 44: Ventilation in the idealized airway block with the needle placed in the far right
position in the trachea at depth 7. Bimodal ventilation is seen here..................81


1
CHAPTER I
INTRODUCTION
The lung is where gas exchange occurs in the body. In order to maximize gas exchange, the anatomy of the lung maximizes the surface area of the air-blood interface in a region of the lung known as the parenchyma where gas exchange occurs. The parenchyma is made up of fragile, microscopic sacs that fill with air, called alveoli [1, 2] and the walls of the alveoli contain a dense network of capillaries. Miniscule airways known as the alveolar ducts supply air to the alveoli, and the respiratory bronchioles supply air to the alveolar ducts. They receive air from the two bronchi, which bifurcate from the trachea - the primary air supply lumen for the lungs. At the top of the trachea, there is a muscular section referred to as the larynx, that includes two folds called the laryngeal folds, which protect the airway, and provide phonation, or speech [3,
4]. Figure 1 shows a diagram of the anatomy of the human lung.


2
Bronchi, Bronchial Tree, and Lungs
Larynx —► *']
Primary bronchi
Secondary bronchi
Tertiary bronchi
Bronchioles
Cardiac notch
Pulmonary vein
Pulmonary artery — Trachea
Alveolar duct Alveoli
Figure 1: The anatomy of the human lung [5]
Respiration is driven by the diaphragm, a dome shaped muscle located at the base of the lungs, and the intercostal muscles, which run between the ribs. When the diaphragm contracts, is pulls down on the lungs and the intercostal muscles expand the chest wall to reduce the plural pressure [2, 6]. This negative pressure pulls the lungs outward, lowering alveolar pressure and drawing in ambient air to fill the lungs. Pressure in airways during normal, relaxed breathing will not greatly exceed atmospheric pressure, and due to the mechanics of respiration there is no way for natural breathing to fill the lungs beyond capacity.
During surgery, anesthesia is used to prevent pain and discomfort in the patient. Drugs
used in anesthesia include analgesics, sedatives, and paralytics, which can all reduce or prevent


3
normal respiration, so gas exchange must be artificially provided via mechanical ventilation [7]. The airway is typically maintained through endotracheal intubation, which provides a path for exchange of oxygen and carbon dioxide and introduction of anesthetic gasses. The endotracheal (ET) tube is sealed in the airway with an inflatable cuff. The ET cuff is a balloon like device surrounding the ET tube that is inflated to expand in the trachea, pushing against the tracheal wall to create a seal and provide a closed ventilation system [8]. Airway pressure can be maintained at appropriate levels and the mixture of inspired gas can be set as necessary for the procedure. Exhaled carbon dioxide levels may be monitored as well to ensure that proper gas exchange is occurring [9].
Modern clinical ventilators use positive pressure to force air into the lungs and, as such, it is possible for the lungs to be filled beyond their natural capacity. In addition, heterogeneous ventilation can, and does, regularly occur during mechanical ventilation [10, 11]. This is when an airway receives excessive or insufficient ventilation relative to the surrounding airways. In the case of excessive ventilation, the airway or alveoli experience large volumes and pressures that can distend and damage the tissue. Insufficient ventilation results in reduced gas exchange between the blood and air in the alveolar sacs, resulting in hypoxemia and hypercapnia. Hypoxemia and hypercapnia can cause complications ranging from temporary organ dysfunctions to death [12-14]. In addition, low airway pressures may result in the collapse, or derecruitment, of small airways and alveoli [15].
Injury caused by ventilation is referred to as ventilator-induced lung injury, or VILI. Atelectatic regions in the lungs will cyclically recruit and derecruit resulting in atelectrauma.


4
Atelectrauma is caused by a repeated fluid-mechanical stress on the alveolar walls that cause direct injury to the epithelium of alveoli and small airways [16, 17]. Injury caused by overdistention of the alveoli is termed volutrauma. Because alveoli share septal walls with their adjacent neighbors there is a mechanical interdependence between alveoli. This may potentiate volutrauma during heterogeneous ventilation where inflated alveoli are adjacent to poorly aerated alveoli [18, 19]. The goals of preventing volutrauma and atelectrauma frequently require conflicting ventilation settings. As such, ventilation is a balancing act of maintaining high enough pressure and tidal volume to maintain gas exchange, and high enough PEEP to prevent atelectrauma, while keeping pressure low enough to avoid volutrauma [20-23]. Furthermore, VILI forms its own positive feedback loop. If an airway is collapsed or flooded, it will promote further injury. As VILI progresses, recruitment and decruitment occur more frequently at higher pressures promoting further atelectrauma [24]. Flooded or collapsed sections of the lung promote heterogeneous ventilation in the dependent airways, further reducing the homogeneity of ventilation, and promoting more damage [25]. As damage accrues in the lung, gas exchange is reduced and more aggressive ventilation is required to prevent hypercapnia and hypoxia.
Endotracheal intubation cannot be used in upper airway procedures such as subglottic and tracheal stenosis surgeries, rigid bronchoscopies, airway granuloma removals, and many other procedures where airway access is necessary. Intubation blocks access to the airway and an adequate seal from the ET cuff may not be possible if the integrity of the tracheal wall is
compromised during the procedure. An open ventilation system must be used, typically in the


5
form of jet ventilation. During LFJV, a needle connected to a high-pressure supply line is typically placed in an endoscopic device, such as a laryngoscope or a subglottiscope. Laryngoscopes and subglottiscopes are tools used to hold the airway open and allow for surgical and diagnostic tools to be placed in the airway. The subglottiscope allows for deeper placement into the airway than the laryngoscope if needed or preferred by the surgeon. All forms of jet ventilation deliver a jet of high-pressure air down the trachea and ambient air is entrained in the flow.
Figure 2: Air entrainment in a laryngoscope
Entrainment is a hydrodynamic property of fluids to carry, or entrain, surrounding fluids into their flow [26]. In the case of jet ventilation, the high velocity, turbulent jet creates a low pressure area, similar to the Bernoulli effect, and ambient air is pulled into the path of the jet [27, 28]. This is shown in Figure 2. This means that distal airway pressure is difficult to control because the volume of air delivered is a function of the supplied air, ambient airthat is pulled down the trachea by the jet, and air that flows back out the trachea and past the incoming jet.
The total flow from a laryngoscope including entrainment has been measured at 20 times the


6
flow produced by the jet [27]. The lack of control makes consistent ventilation difficult and may increase the occurrence of VILI. Methods to provide safe, homogeneous ventilation via jet ventilation have been insufficiently studied. Our long-term goal is to make pressure adjustments for LFJV more of a science and less of an art. We postulate that evidence-based written standards for LFJV and a device to produce repeatable flow readings will help to optimize practices and reduce associated guesswork for anesthesiologists using the technique, improving patient outcomes.


7
CHAPTER II BACKGROUND
Low frequency jet ventilation (LFJV) is a widely used form of ventilation. It is frequently used during rigid bronchoscopies, emergency ventilation in 'can't intubate, can't ventilate' scenarios, and surgery for laryngotracheal stenosis where intubation is not possible or would obscure the surgical field [29-31]. In rigid bronchoscopy, LFJV is currently the most commonly used form of ventilation [30]. While statistics on the number of LFJV cases per year are not readily available, based on the frequency of use in the procedures listed above, we estimate there are 200,000 LFJV cases in the US per year. Pressure adjustment for jet ventilation is currently performed empirically using a low precision pressure regulator, shown in Figure 3. The pressure is then tested by applying air from the ventilation needle onto the
anesthesiologist's forearm [27].


8
Figure 3: A standard jet ventilation setup. The red arrow is pointing to the pressure regulator.
While LFJV remains in widespread use, there is limited data on settings that optimize oxygenation and lung protection using this technique. Without continuous capnography provided by a closed anesthetic circuit, it is not possible to accurately measure end-tidal CO2 concentrations in real time and thus adequacy of ventilation is based on intermittent blood gas concentration sampling [28]. There is currently no written standard for low frequency jet ventilation pressures and flow rates, leaving anesthesiologists to rely on personal experience and training rather than evidence-based guidelines. As such, methods to set pressure vary by provider and institution. Many set by feeling for the correct pressure on their forearm while others start at a low pressure and then slowly increase pressure during ventilation until they see adequate chest rise [27, 32]. The University of California - San Francisco recommends


9
starting with a flat setting of 50 PSI for adult patients [33]. Applying insufficient pressure can result in hypoventilation leading to hypoxia and hypercapnia which can cause organ failure, permanent brain damage, and in extreme cases, death [12-14]. Applying excess pressure can result in volutrauma while insufficient PEEP can cause atelectrauma [16,17]. The lack of data-based standards and hardware to allow repeatable jet settings leads many providers to avoid jet ventilation altogether even when it may provide the optimal form of ventilation for a given procedure. Moreover, the use of qualitative jet settings makes it difficult to determine and share effective parameters between providers and institutions.
Low frequency jet ventilation was introduced in the 1950s as a method of non-intrusive ventilation during upper airway surgery [34]. High frequency jet ventilation, another form of jet ventilation, was developed shortly after and uses low tidal volumes at high respiratory rates of 100-600 breaths/min [35, 36]. The first measurement of jet ventilation pressures was over 20 years later, in 1977, when tracheal pressure was measured and it was determined that high frequency jet ventilation could generate small amounts of positive end expiatory pressure (PEEP) [37]. Recommended safe jet inlet pressures range from 20 to 50 psi depending on the institution [32, 33, 36]. Note that these inlet pressures are three to four orders of magnitude above the distal airway pressure.
Like other forms of ventilation, there are complications associated with LFJV including hypotension, hypertension, hypercapnia, and hypoxemia [38]. Studies of LFJV during rigid bronchoscopy found that oxygenation can be maintained at adequate levels during ventilation while barotrauma-associated complications, such as cervical emphysema and tension


10
pneumothorax, remain under 1% [30]. Another study showed that jet ventilation is an effective method of rescue ventilation during operations for throat cancers. That study established advantages in decreased surgical obstruction and increased gas exchange compared to rescue by mask ventilation by successfully restoring proper blood oxygen levels in all 31 patients included in the study [39]. For ductus arteriosus ligation, there was no significant difference in gas exchange, or rate of complications between jet ventilation and other forms of ventilation [40],
Jet ventilation is currently the gold standard for 'can't intubate can't ventilate' situations but the rates of barotrauma and total complications may be as high as 32% and 51% respectively, even in situations where jet ventilation is considered the best method [31]. That a literature review that noted how frequently other sources listed barotrauma as a complication and specifically included pneumothorax, pneumomediastinum, and subcutaneous emphysema as complications. Note that these barotrauma rates are much higher than reported during rigid bronchoscopy and this may be due to the challenging situation where emergency ventilation is applied. Obesity has been shown to have little to no correlation with complications during jet ventilation [41-43]. This is a somewhat surprising result given that the increased weight of the chest in obese subjects is known to reduce respiratory system compliance, and thus tidal volumes, at a given pressure. Although there have been a number of studies of LFJV, none have directly addressed what is possibly the most important question: how can the technique be applied in a consistently safe and effective manner?
As such, anesthesiologists still have no evidence-based, quantitative guidelines for safe


11
pressure or flow rate settings. Furthermore, the jet ventilation pressures are set according to the inlet pressure supplied to the needle and by feeling the jet of air coming out of the needle [44]. Since the flow rate delivered to the patient also depends on equipment downstream of the pressure regulator, we assert that this is not a suitable approach. These are major shortcoming since improper pressures or techniques may lead to heterogeneous ventilation, barotrauma, impaired gas exchange, and pneumothorax [45].


12
CHAPTER III
MATERIALS AND METHODS
We designed airway models to explore how jet pressure, flow rate, needle placement in the trachea, and laryngoscope and subglottiscope use affected distal airway pressure. These models were fitted with pressure transducers and then circuitry and software was created to record pressure data. A flow measurement device was developed to quantify flow rate including flow entrainment. Our experimental setup was designed to characterize the fluid dynamic properties seen in human airways relevant to LFJV.
Model Airway Pressure Measurements
The experiment uses two different physical models of the upper airway. The first model is an idealized representation of the first five generations of the human airway tree and is referred to as the 'Idealized Airway Block'. The design is unique to this experiment and the branch angle, size, and generational restrictions are derived from works by Florens, Horsfield, and Weibel et al. [46-48]. Our model is symmetric in the sagittal and coronal planes. The first generation is a 125mm long and 18mm diameter trachea. Each of the four bifurcations has a 70-degree branch angle and the plane of each bifurcation is rotated perpendicular to the previous bifurcation about the axis of the superior duct. The diameter at each generation of the airway was 79% of the size of diameter of the superior airway. Each generation, except for the trachea, maintained a diameter to length ratio of 1:3. This idealized airway model allows
the distal airway pressure relationship to be studied purely as a function of needle placement in


13
the trachea. The symmetry in the block also allows for verification of results by checking that symmetrical ventilation produces symmetrical results.
Figure 4 : Idealized airway geometry based on the morphometry models from Florens,
Horsfield, and Weibel [46-48].
We also considered a full-size fabrication based on a CT scan from a healthy 17-year-old male human subject, shown in Figure 5. This model, referred to as the 'CT Airway Block', was originally developed for bronchoscopy training and was kindly donated by Emily DeBoer and Jennifer Wagner of the University of Colorado. The length of the trachea was 129mm.


14
Figure 5: Model of the human airway tree, derived from the CT scan of a 17-year-old healthy
human subject.
The idealized and CT airway models are each comprised of a solid block of 50 durometer silicone with a hollow airway inside. The idealized airway was designed in SolidWorks, and the CT scan was imported. Positive casts were then 3D printed out of calcium sulfate hemihydrate with a Projet GP 660Pro (3DSystem, USA) with and surface-coated with paraffin wax. The print was encased in silicone and then dissolved using a heated ultrasonic bath, creating a hollow airway in the silicone block. Holes for access ports were drilled in the blocks to reach the fifth airway generation in each lobe of the lung in the CT airway model, and each of the sixteen
distal airways for the idealized airway model. Silicone adhesive was used to mount barbs at the


15
ports to allow sensors to be easily attached and removed. This created a sealed airway that allowed for easy measurement in the distal airways.
Jet Needle Positioning System
In order to investigate the effects of different jet positions in the trachea, an automated system was designed and fabricated to control the position of the ventilation needle. An Anet A8 3D printer with a precision of 0.2mm was used as the base of the system and a fixture, shown in Figure 6, was designed to hold the lung cast in place on the printing platform. The fixture was bolted to the platform of the 3D printer and wheels were attached to allow for smooth movement. The main printing code was deleted from the printer and G-code was written for movements needed for ventilation. A fixture was designed to hold a Luer Lock ventilation needle where the printer extrusion head was originally located. For this experiment, a 100 mm 14-gauge needle was used, as is typical for jet ventilation. The mount was designed to hold the needle level and pointing straight ahead for maximum stability and repeatability. The gas delivery system was configured with an electric solenoid valve and a manual button in parallel to allow for electrically actuated or manually triggered ventilation. The electric valve was connected to the printer control system so that ventilation experiments were fully
automated. The system is shown with the CT airway block in Figure 6.


16
Figure 6: Photo of the experimental setup with the CT airway block. The regulator is indicated
with a red arrow.
The system was tested for consistency and can make repeatable maneuvers. Test measurements of repeated runs in the same position had consistent distal airway pressures with a standard deviation of less than a percent difference. The blocks were highly sensitive to changes in position. In one case, a misalignment of the block on the ventilation platform caused the block to be rotated about the trachea by about 3 degrees. This resulted in substantial changes in distal airway pressure. In several cases, a 2mm difference in needle positioning resulted in a change of distal airway pressures of over 50%. Consistent reading of distal airway pressures from repeated runs indicates consistent needle positioning. A rigid tube was fitted to the trachea for each model so that the needle could be aligned next to the tube to
verify that the needle was correctly aligned to be concentric with the trachea. It was measured


17
to be level with the surface of the desk, the printing platform, and the walls of the printer. The printer could repeatedly place the needle directly next to the alignment tube within a small fraction of a millimeter and parallel. We verified that the needle stayed parallel in directly next to the alignment tube as the needle moved towards and away from the tracheal opening.
Signal and Data Processing
The system measures distal airway pressures within the airway block as a function of jet needle location in the trachea (dorsal-ventral and left-right) and depth into the trachea. For each position, the system is given a second of dwell time to allow residual vibration from moving to slow or stop before ventilation is engaged for three seconds. A MATLAB program was written to show live views of the data, make recordings, demultiplex the recorded signals, plot the data, filter noise, and analyze the resulting data. Signal analysis included the mean pressures at each sensor for each depth, standard deviation, variance, homogeneity across sensors, and homogeneity across depths. A GUI was written to display data and conveniently export plots or excel data.
We selected seventeen Honeywell transducers to record our pressures. We used HSCDANN001PGAA5 1 psi high-precision analog gauge pressure transducers and SCDANN001PDAA5 1 psi high-precision analog differential pressure transducers to measure distal airway pressure. A Honeywell HSCDANN060PGAA5 60 psi high-precision analog gauge pressure transducer was used to measure the pressure in the supply line. A 1.5 cm segment of 1/4" silicone tube was affixed to each sensor to allow sensor to be easily attached and removed
for calibration. The output signal from each passed through a Microchip Technology MCP6282-


18
E/P-ND op-amp to act as a buffer to reduce noise. The output from the op-amp is routed through an RC low pass filter set to 210 Hz. This frequency was set as an anti-aliasing filter before the data was recorded by the DAQ. The sensors were connected to a National Instruments USB-6009 eight channel, 14-bit data acquisition system (DAQ) which was then routed to a computer.


Figure 7: Wiring diagram for the data acquisition system from the pressure transducers to
computer.


20
A 4:1 multiplexer was used to increase the number of signals that could be acquired with the DAQ. A two-bit binary counter was used to cycle the multiplexing chips. During each data acquisition sequence, 17 pressure signals and 2 binary counting signals were measured. The 17 pressure measurements include the 16 distal airway pressure sensors and the inlet pressure. The binary count signals were continuously acquired on their own channels to allow sensor identification during demultiplexing. Each of the 16 airway pressure signals was acquired at 6.144 kHz for 1 ms out of every 4 ms (25.152 kHz asynchronous sampling for all DAQ input channels). The inlet pressure transducer had its own channel for uninterrupted data acquisition. The analog circuit used an Nexperia USA Inc. 74HC4052D,653 multiplexer chip and a 555 timer (960 Hz) connected to a Texas Instruments SN74LV393APWR four-bit counter, set as a two-bit counter to control the switching. A 555 timer was set to run in a self-triggering, bistable mode and was set to 960 Hz. This 960 Hz binary count switched the MUX between four sensor inputs giving a final effective sampling rate of 240 sampling sets per distal airway sensor, per second or approximately 1500 samples per channel, per second.
The MUX cycled acquisition between the 4 sensors on each channel and, for each sensor, a sampling set was recorded that contained six to seven data points over the course of approximately one millisecond, followed by three milliseconds of no data for the sensor.
During the demultiplexing process, the software used the binary data to assign input data from each channel to the corresponding pressure signal. The data from the multiplexed signals would occasionally be recognized as part of the wrong channel because the asynchronous reading from the DAQ caused off timing between the binary count channels and the data


21
readings. An off-count data point would show up as a spike containing a single data point. The data was processed with a 10th order median filter to remove the spikes. A 9th order MR 200 Hz low pass filter was also applied to the pressure data.
Airway Pressure Measurement Procedure
The automated test apparatus ventilated the airway tree in thirteen different positions in the tracheal cross-section. Each position in the trachea was considered a different test run and data were recorded at 7 different depths during each run. Figure 8 shows the thirteen different positions in the tracheal cross-section that were tested. The block was ventilated in the center of the trachea and three steps in each direction. The outward steps ended with the center of the needle 2mm from the wall of the trachea. This provided a gap of just under 1 mm between the tracheal wall and the outside wall of the needle. The step spacing leading to the wall was set to 2 mm. For both the CT and idealized airway blocks we chose to use a uniform step distance and the same final distance from the wall rather than evenly spaced steps. The two tracheas had somewhat different shapes and diameters and we hypothesized that fluid dynamic surface effects at the jet-tracheal wall interface would affect homogeneity and would
be a function of distance from the tracheal wall.


22
Ventral
Figure 8: A top-down view of the trachea with the thirteen ventilation positions shown: The
center, and three steps in each direction.
Flow Measurement
We developed a venturi flow measurement device (Figure 9) to allow the jet ventilation system to be set to a specific flow rate. Most jet ventilation systems start by setting the pressure in the supply line before the needle but needle gauge, length, and dynamic losses (e.g. friction in the tube, viscous losses from an elbow) can substantively change the flow rate
produced at the same pressure.


23
Ai
010
Figure 9: Venturi flow meter design. The locations of Ai, kj, Pi, and P2 (Eq. 1) are labeled in the
cutaway (A).
Most ventilation flow meters currently on the market measure closed system flow which is relatively trivial. Jet ventilation is an open system, causing outside air to be entrained into the jet flow and pushed into the airways. This means that the total flow may be many times higher than the flow coming out of the jet needle. One study noted a 20-fold increase in
flow rate from a jet inserted in a laryngoscope [27]. This entrained air substantially changes the


24
flow dynamics inside the trachea during ventilation and must be considered for an accurate reading. The amount of air entrained is based on several factors which are controlled by the needle mounting system shown in Figure 10.
Figure 10: Jet needle mounting system
In order to record repeatable measurements, the jet needle must be concentric and parallel to the venturi tube, stable enough to avoid vibrational or oscillatory effects in the jet, and flush with the entrance of the tube. Our system uses a mount to hold the needle concentric and parallel to the venturi flow meter and the mounting system is rigid, to prevent
vibration or movement in the needle. 3D printing was used to manufacture the venturi flow


25
meter. To smooth out the texture in the printed plastic, an acetone vapor bath was constructed. The print was placed in an acetone vapor filled chamber for approximately 30 minutes and allowed to sit while the acetone softened the plastic. The acetone source (paper towel damp with acetone) was removed and the chamber was opened and allowed to clear. The print was given approximately an hour to re-harden before it was removed from the chamber.
The venturi flow meter requires a hydrodynamically fully developed flow to function correctly. The jet produces a highly turbulent, undeveloped flow which requires either a long straight section or a flow conditioner to produce a fully developed flow before reaching the converging section of the tube. We used a long entrance length to develop the flow. The flow was calculated from the measured pressure difference at ports Pi and P2 (Figure 9) using the Bernoulli equation
where flow (Q) is a function of the cross sectional area of the entrance (Ai), the cross sectional area of the constriction point (A2), the pressure before the entrance (pi), the pressure at the constriction (P2), and the air density (p). The curve for the flow rate compared to the measured
Q-A
P1-P2 . 2
(1)
P
voltage is shown in Figure 11.


26
Calibrated Bernouli Equation
Figure 11: The curve of the flow rate vs the measured voltage. This provides an accurate measurement from a flow rate of 0.25 L/s to approximately 3.25 L/s.
Calibration of the meter was conducted with a calibrated 3 L syringe. The flow meter was connected to the DAQ to record Q as 3 L of air was pushed through the meter at different flow rates. The integral of the flow rate during the recording was the total flow through the meter and should be 3 L for each test.
Additional Considerations
Our mounting system is capable of holding multiple different needle gauges and lengths
and accepts the standard needle mount bracket that will be used to subsequently mount the


27
needle in the laryngoscope. We also designed a stable, twist-lock mounting bracket shown in Figure 12. This allows the bracket to be quickly and easily mounted in stable and repeatable fashion.
Figure 12: Twist on Mounting Bracket
The needle mount is easily removed from the device and could be made as a plastic disposable piece or metal reusable part that can be sterilized in an autoclave. This is necessary because the mount will be in direct contact with the ventilation needle that will be inside the patient and therefore must be sterile. Our system includes a large, 16-pin, 5-volt, 16x2 backlit LCD display module that gives the real time flow rate as well as an average flow rate over the last several seconds of reading. We used an Arduino to run the code and a custom built circuit
to amplify and filter the sensor data from the pressure transducer. It displays a live view of the


28
instantaneous flow rate and a running average of the flow for the three preceding seconds. Figure 13 shows our functioning prototype of the device without the housing and with a simplified mounting bracket.
Figure 13: Functioning prototype of the air flow entrainment measuring device, with housing
removed


29
CHAPTER IV RESULTS
We first consider jet flowmeter characterization and accuracy. The effect of needle position in the trachea is explored and relationships between inlet pressure, flow rate, and distal airway pressure are defined. Finally, the effects of a laryngoscope and subglottiscope on the distal airway pressure are described.
Flowmeter Characterization
15 flow rates were tested ranging from 0.24 L/s to 2.3 L/s and the meter showed good accuracy across the range tested (Figure 14). At flow rates above 0.7 L/s the mean measured volume was 2.99 liters with a standard deviation of 0.02 liters or 0.7% of the total volume. The mean error was 0.006 liters or 0.2%. The flowmeter was less accurate for flow rates below 0.7 L/s, reading an average volume of 3.14 liters with a standard deviation of 0.037 liters for a mean error of 4.7% with a standard deviation of 1.2%. Since the flowmeter was accurate at flow rates above 0.7 L/s but read consistently high at low flow rates a nonlinear calibration curve could be used to further improve accuracy. However, this was not necessary since our experiments, and clinical jet ventilation, use flow rates above 1.2 L/s and this places the
measurements in the high-accuracy range of the meter.


30
3.2 r
3.15
O
O
O
O
(D
E 3.1
o
>
~CJ
s^_
S 3.05 cti
0J
O
3 -
2.95
O O
O
O
O
0.5
O
O
O
1 1.5 2 2.5
Flow (L/s)
Figure 14: Volume determined as the integral of the flow measured using the venturi flowmeter (blue circles). The target value for each measurement is 3 L (red line) that is delivered using a
calibrated syringe.
Airway Pressure Measurements
Each experimental run ventilated the airway block at one position in the tracheal cross-section and seven depths into the trachea. For each depth, data was simultaneously recorded at 16 different distal airways positions. Figure 15 shows the raw voltage reading from the 17 sensors with the jet needle 6 mm from the left wall of the trachea in the CT airway block. The first ventilation depth (Dl) is 1 cm past the location of the vocal folds in the human subject, and each step was one centimeter deeper into the trachea. The final depth (D7) is 5.9 cm above the carina. In the idealized airway block, D7 is 5.5 cm above the carina. The low voltage (and low


31
pressure) times are when the apparatus is repositioning the jet needle. The high voltage (and high pressure) times are when the jet ventilation is applied. Note that the voltage-pressure relationship is different for the inlet pressure (black), channels 1-8, and channels 9-16.
Figure 15: The demultiplexed, unfiltered, uncalibrated voltage readings from each sensor during a ventilation test run on the CT airway block. The needle was centered and moved through all 7 depths. D1-D7 denote the depth into the trachea from 1cm to 7cm.
Sensor-specific calibration curves were used to convert the raw voltage readings to the pressure readings shown in Figure 16a. Each line represents the demultiplexed, median-filtered pressure at one distal airway. These data are summarized in Figure 16b where the mean distal airway pressure across all sensors is shown with a heavy line and the range from minimum to maximum pressure across all sensors is shown with a shaded area. This shows the homogeneity of ventilation at a given depth and radial position in the tracheal cross-section. Areas where the shaded range is small indicate homogeneous ventilation while positions with


32
heterogeneous ventilation show wide spreads in the shaded area.


33
Figure 16: Distal airway pressure measurements in the CT airway block with the jet needle 6mm from the left wall of the trachea (first step to the left) showing the filtered pressure for all 16 distal airway sensors (a). The homogeneity graph (b) shows the mean pressure (heavy black line) and the shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. Note that (b) is normalized by the inlet pressure. D1-D7 indicate depth into the trachea from 1cm to 7cm.
Data Normalization
Distal airway pressures were divided by inlet pressures to define the normalized pressure, shown in equation 2, where Paw(n) is the pressure in airway (n) and Pin is the inlet pressure. This scaling was performed because the inlet pressure varied over time. Given these variations, it would seem intuitive to find a more consistent regulator. However, we elected to keep the regulator because it was substantially similar to the actual regulator used in surgical ventilation. The pressure regulator provides an initial spike in pressure and then decays to less than 90 percent of the peak value over a few seconds as shown in Figure 17. The initial experimental configuration had additional pressure transients due to the uneven pressure provided by the compressor. That artifact was eliminated by the addition of a 10-gallon auxiliary air tank in series with the compressor.
Pnorm ~ Paw (^0 /Pit
(2)


34
Figure 17: Transient in inlet pressure over the three second ventilation burst provided during the experiment. This is measured by a pressure transducer placed immediately before the
ventilation needle.
There were irregularities in the inlet pressure primarily from the regulator, however there was also a long-timescale pressure decay that we attribute to changes in auxiliary tank pressure and short-timescale transients that we attribute to inlet tubing compliance. These transient values in the pressure supplied to the ventilation needle were reflected in the flow rate and therefore the distal airway pressure, as they would in an actual jet ventilation setup.
The data acquisition system had a continuous reading from the sensor recording the pressure at the ventilation needle. This reading was used as the reference for normalizing the
data. By dividing the distal airway pressures by the inlet pressure, the effect of these transients


35
were reduced to facilitate further analysis of the measurements
Airway Pressure Relationships
One of our primary interests was determining the distal airway pressure as a function of inlet pressure or flow, and several equations were derived to relate the values. Since we have a consistent ventilation setup, any given inlet pressure correlates directly to a single flow rate and therefore we can translate inlet pressure to flow rate using the Darcy-Weisbach equation. For more information on the flow measurement device, see the device design considerations in §3.3. We were consequently able to create equations relating both inlet pressure and flow to distal airway pressure. Figure 18 shows the relationship between total flow rate and distal airway pressure. This is consistent across different depths, as shown in Figure 19.


5.01SSX^ -4.8206X+2.7803
36
0)
c-
co
ii
0>
_D
ra
>
oo
00
(JD
s
Q.
Q
*
O
CQ
00
CO
CO
c-
c\i
3
c-
O
00
X
9
0)
c-
co
ii
0)
ns
>
co CM i\D
* s
Tf O
CM
IN O II
^ U0 I. *
T3 ^r CL .C
CL
0)
Q
t
t
q
(J
5 5 s
II
0)
_3
(0
>
CO
in
If)
s
CM
g
CM o o
in â– 
X
O)
U0
S .1
in .. a
13 Is-
CL £
4->
Cl
oj
Q
CO
o
( Figure 18: Relationship between distal airway pressure and total flow rate including entrained air. This is in the center position of the idealized airway block at 1 (a), 4 (b), and 7 cm (c) into
the trachea.
Flow Rate (L/s)


37
Figure 19: Relationship between distal airway pressure and total flow rate shown at 1, 4, and 7 cm into the trachea, with the needle centered in the idealized airway block.
Because clinical jet ventilation is guided by inlet pressure we have performed the remainder of the analysis using that parameter. Figure 20 shows the strong linear correlation between inlet pressure and distal airway pressure for the idealized airway block.


38
Figure 20: Distal airway pressure vs inlet pressure for the center needle placement in the idealized airway model at depths of 1 (a), 4 (b), and 7 cm (c).
Figure 21 and Figure 22 show the inlet pressure versus distal airway pressure for the idealized and CT airway models, respectively. In both cases there is a strong linear correlation
between inlet and distal airway pressure.


39
Figure 21: The linear regression of inlet pressure vs distal airway pressure for the CT airway model with the needle centered in the trachea at depth 1.
The CT airway model has a linear relationship between distal airway pressure (Pdistai) and inlet pressure (Pin) (Eq. 3) and R2 = 0.9977 as shown in Figure 21. The best fit to these data suggests that a distal airway pressure of 0 is achieved when the inlet pressure is 500 cmbhO. However, at pressures below approximately 700 CIT1H2O there was a nonlinear section that tapered off and a distal airway pressure of 0 was measured at an inlet pressure of 30 CIT1H2O. A figure is not included as this is well below the pressure used in ventilation. As such, Eq. 3 should only be used for inlet pressures above 700 CIT1H2O.
Pdistal = 0- 0064 * Pjei — 3.19 (3)


40
Figure 22: The linear regression of inlet vs distal pressure for the idealized airway model with
the needle centered in the trachea at depth 1.
For the idealized airway model, there was a linear relationship between Pdistai and Pin (Eq. 4) with R2 = 0.9985 (Figure 22). Again, there was a nonlinear section for Pin < 250 cmFhO and equation 4 should not be applied below this threshold.
Pdistal = 0- 0065 * Pjet — 1.25 (4)
Airway Pressure Distribution
Homogeneity grids were created to describe the distribution of distal airway pressures.
Figure 23 shows the numbered locations of the sensors on the idealized airway block, shown


41
from the inferior end of the lung block looking up. Figure 24 shows the corresponding homogeneity grid, which represents a view from the superior end of the block looking down the trachea.
Figure 23: The numbered positions of the sensors on the idealized airway block


42
1 5 9 13
2 6 10 14
3 7 11 15
4 8 12 16
Figure 24: Sensor represented on the homogeneity grid.
The bifurcations in the idealized airway tree rotate 90 degrees each generation. The first-generation bifurcation is left-right, the second-generation bifurcation is anterior-posterior, and so forth. For example, sensor 9 represents the airway that goes right at the first bifurcation, anterior at the second bifurcation, left at the third bifurcation, and up at the fourth
bifurcation, as shown in Figure 25.


43
Figure 25: Example of airway bifurcations to reach sensor 9. Each darker shade represents a
branch farther into the block.
Figure 26 shows homogeneity grids for the needle centered in the trachea at depths one, four, and seven for the idealized airway model. With the needle centered, the idealized airway model did not show a substantial change in pressure with respect to depth of needle placement. At depth seven (5.5 cm from the carina), even a millimeter of movement towards one of the tracheal walls could result in heterogeneous ventilation (data not shown).


44
Pressure Homogeneity at Depth 1; Center
a[
B
Pressure Homogeneity at Depth 4; Center
x10
-3
6
5.5
Pressure Homogeneity at Depth 7; Center
x 10“'
Figure 26: Comparison of homogeneity in the distal airways for the idealized airway model at
depths 1 (a), 4 (b), and 7 cm (c).
For the CT airway model, the trachea was curved, the diameter was variable, and it had an irregular cross-section that changed with depth. This made it much more difficult to determine what the center of the trachea was, especially for a system designed to move in a straight line through the trachea. The airway homogeneity was sensitive enough to the position
Normalized Pressure (Ten Thousandths of Inlet Pressure}


45
of the ventilation needle that a few millimeters of difference from the true center would make a clinically relevant impact in homogeneity. The normalized pressure distribution shown in Figure 27 was measured with the jet needle centered in the tracheal opening and downward in a direction parallel to the tracheal centerline.
Figure 27: The homogeneity grids at depths 1 (a), 4 (b), and 7 cm (c) with the jet needle centered in the trachea in the CT airway block.
Normalized Pressure (Ten Thousandths of Inlet Pressure}


46
The CT airway block (Figure 27) shows a marked decrease in distal airway pressure and increased pressure heterogeneity as depth into the trachea increased. This is in contrast to the idealized airway model which showed only modest increased in heterogeneity and pressure with increased depth. With the needle situated closer to the tracheal wall (Figure 28), the pressure was more likely to favor a specific bifurcation, and the model typically showed bimodal ventilation where one region had higher pressures than the other. Trimodal ventilation was observed in the CT airway block (Figure 28c) when the needle was positioned along the right wall of the trachea at a depth of 7 cm. The same position for the idealized airway model is available in the appendix for comparison.


Figure 28: Trimodal ventilation demonstrated in the CT airway block with the needle placed the far right position in the trachea at depths of 1 (a), 4 (b), and 7 cm (c).
Figure 29: The homogeneity grid at depth 1 with an 18 gauge needle in the center of the
trachea for the CT airway tree.


48
The diameter of the jet needle alters the flow rate and, as such, has a profound effect on distal airway pressure. Figure 29 shows the homogeneity grid for the CT airway block with the needle centered in the trachea at a depth of 1 cm. However, in this case the 14-gauge needle was replaced with an 18-gauge needle. With the same inlet pressure setting, changing the needle from 14 (Figure 27) to 18 gauge (Figure 29) reduced the flow rate by 50.4% from 1.25 L/s to 0.62 L/s. This resulted in a normalized distal airway pressure decrease of 65.7% which is in accordance with the relationship shown in Figure 19.
In nearly every ventilation case in both airway models the ventilation pressure in the distal airways became more heterogeneous as the ventilation needle was moved farther into the trachea and/or closer to the tracheal wall. The sensitivity to distance from the tracheal wall increased with depth into the trachea. This is likely due to the turbulent, undeveloped flow produced by the jet ventilation needle. The trachea, acting as a tube entrance length, will produce a fully developed flow regime given enough length to do so, and a longer entrance length will more effectively develop the flow. To demonstrate these trends, homogeneity graphs were created as outlined earlier in this chapter (e.g. Figure 16). Figure 30 below shows the homogeneity of ventilation in the idealized airway model at varied distances from the left
side of the trachea.


49
Depth into Lung Cast (cm)
Figure 30: Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left side of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest
to the tracheal wall.
The idealized airway model did not show a decrease in pressure with increased depth into the trachea for the center or the first step to the left. A decrease in pressure with depth was present but not continuous for left 2 and there was a small but continuous decrease in pressure for left 3. Every needle placement except for center showed a decrease in homogeneity with increased depth and this effect was more pronounced as the needle moved
closer to the tracheal wall. At a depth of 1 cm past the laryngeal folds, the homogeneity was


50
similar for all distances from the wall and there was a difference of approximately 10 percent in the mean pressure between the highest and lowest value. As depth increased, there was a substantial difference in pressure homogeneity and mean pressure with placements closer to the tracheal wall.
5.5
5
=3
dj
Q_
U 4.5
_N
“ra
E
o
4
3.5
01234567B Depth into Lung Cast (cm)
Figure 31: Comparison of needle placements in the idealized airway block for the center and the placements closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.
x10'3
T---------------------1--------------------1---------------------1--------------------1---------------------1--------------------T
\
Center
Ventral
Dorsal
Left
Right
Figure 31 shows the homogeneity graph comparing the center placement and the step
closest to the tracheal wall in each direction. For all movement directions, the pressure was


51
relatively homogeneous for proximal needle positions and the normalized pressure were approximately equal. At more distal needle positions homogeneity decreased and the homogeneity of distal airway pressure became more sensitive to small changes in needle placement.
Figure 32: Homogeneity in the distal airways of the CT airway block as a function of distance from the left side of the trachea. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wall.
In the CT airway block pressure and homogeneity tended to decrease with increasing jet
needle depth (Figure 32). This trend was seen in the left and right placement and, to a lesser


52
extent, in the ventral and dorsal placement.
Figure 33: Comparison of needle placements in the CT airway block for the center and the placements closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.
Figure 33 shows the CT airway block with the needle placements closest to the tracheal wall in each direction, as well as the center placement. The homogeneity consistently decreased with increasing depth into the trachea. The pattern of pressure loss with distal placement of the ventilation needle seen in Figure 32 is vaguely present but not nearly as
defined.


53
Laryngoscopes and Distal Airway Pressure
Laryngoscopes and subglottiscopes are used during upper airway procedures to hold the airway open and allow unobscured access for surgical tools. Since these surgical scopes are common and can greatly increase the air entrainment, it is important to study the effects they have on distal airway pressure. An Ossoff-Pilling 230 mm adult male subglottiscope (Model 52245) was tested at depths of 1 cm and 3 cm (Figure 34) below the vocal folds in the CT airway block using a standard Pilling-CS 14-gauge ventilation needle. The end of the subglottiscope is typically inserted one centimeter past the laryngeal folds and the folds press against the outer walls to create a seal. Our experiment used putty to seal the subglottiscope to the opening of the airway block. The normalized pressure averaged just slightly higher than without the subglottiscope and there was no physiologically relevant difference between scope placement
depths.


54
Subglottiscope Pressure Homogeneity, 1 cm
AT
Subglottiscope Pressure Homogeneity, 3 cm
xIO
-3
B
7
6.5
Figure 34: Pressure homogeneity grids for the subglottiscope with the orifice 1 and 3 centimeters past the laryngeal folds. This is in the CT airway block.
The ventilation homogeneity was also measured for an Ossoff-Pilling adult male 168 mm micro-laryngoscope (model 522191) with the standard Pilling-C8 14-gauge needle. The laryngoscope was placed at a depth of 1 cm past the laryngeal folds and the needle was mounted on the bottom of the right wall with the end 2.5 cm into the laryngoscope as is common in clinical practice. Figure 35 shows the same laryngoscope, placed at the same depth, with the same needle mounting position, with the ventilation needle angled up within the
laryngoscope at 10, 5, 3, and 0 degrees.
Normalized Pressure (Ten Thousandths of Inlet Pressure}


55
Laryngoscope Pressure Homogeneity, 10° Laryngoscope Pressure Homogeneity, 5°
0.016 0.015 0.014 0.013 0.012 0.011 0.01 0.009 0.008
Laryngoscope Pressure Homogeneity, 3° Laryngoscope Pressure Homogeneity, 0°
0.016 0.015 0.014 0.013 0.012 0.011 0.01 0.009 0.008
Figure 35: The homogeneity grids for the Ossoff-Pilling Laryngoscope ventilation. The needle is mounted at the bottom of the right wall and tilted up at angles of 10, 5, 3, and 0 degrees. This
is the CT airway model.
In Figure 33 we saw a trend in the CT airway model of increased pressure as the needle moved superior in the trachea. This trend of superior needle placement and increased flow rate resulting in higher distal airway pressures continued with the addition of the
subglottiscope and laryngoscope. The laryngoscope moved the needle placement even farther
Normalized Pressure (Ten Thousandths of Inlet Pressure) Normalized Pressure (Ten Thousandths of Inlet Pressure)


56
from the lungs, and the mean pressure at every angle was higher than it would be without the laryngoscope. By moving the ventilation needle from a 10-degree angle to a 3-degree angle, there was an 82% increase in distal airway pressure. The experimental needle placement was set by a surgeon in the same manner that it would be placed during an actual ventilation, and the 10-degree change in angles is within what would be typical movement during surgery.


57
CHAPTER V DISCUSSION
The primary goal during mechanical ventilation is to exchange gas to maintain homeostasis without injuring the lung. There are three important pressures in ventilation: the peak inspiratory pressure (Pip), plateau pressure (Ppiat), and the positive end expiratory pressure (PEEP). The peak inspiratory pressure (Pip) is the highest pressure in the lungs during inhalation and is a combination of lung resistance and elastance. The pressure at the end of inspiration is Ppiat and this is predominantly governed by the lung elastic recoil and lung volume. The PEEP is the pressure at the end of exhalation that is used to prevent airway and alveolar collapse. LFJV does not provide a PEEP. The tidal volume delivered to the lungs (Vt) = Cst (Ppiat — PEEP), where Cst is the pulmonary system compliance. Gas delivery is defined by the minute ventilation which is the product of tidal volume and respiratory rate and is the amount of gas received per minute. The inspiratory and expiratory pressure that drive ventilation must be carefully selected to deliver sufficient gas while keeping Ppiat low enough to avoid volutrauma and PEEP high enough to prevent damage via atelectrauma [20-23].
If all regions of the lung are provided with an appropriate range of inspiratory and expiratory pressures, then the alveoli will be properly supplied with fresh gas to prevent hypoxia and hypercapnia. However, high Ppiat can drive excessive volumes of air into the lungs and, in the most severe cases, volutrauma will occur. It is also possible that regions of the lung are subjected to different ranges of inspiratory and expiratory pressures (heterogeneous
ventilation). If the spatial mean distal airway pressure across the lung is at a reasonable level


58
but one airway is receiving excessive pressure, this will result in excessive volume and an increased risk of volutrauma in the parenchyma supplied by that airway. If the spatial mean distal airway pressure is reduced to avoid overdistension in the high-pressure region, then the remainder of the lung may receive insufficient flow resulting in impaired gas exchange. By providing homogeneous ventilation, the spatial maximum of the pressure can be reduced while maintaining a higher mean distal airway pressure, allowing for appropriate ventilation to prevent hypoxia and hypercapnia with a reduced risk of complications in the lungs.
The rate of complications specific to LFJV are not well established and consequently, we do not know the optimal settings. Data and statistics from other types of jet ventilation can be used as a general reference for what we might expect to see in LFJV, but should not be considered the same as actual data on LFJV. In high frequency jet ventilation, complications involving gas exchange, including hypoxemia and hypercapnia, occur in 18 and 22% of adult bronchoscopy jet ventilation cases respectively and serious complications including barotrauma, cervical emphysema, and tension pneumothorax occur in approximately 1% of adult bronchoscopy cases [49-51]. In 'can't intubate, can't ventilate' emergencies, a manual transtracheal jet is usually hastily set up and complications are very common, with a barotrauma rate of 32% of cases and a total complication rate, including hypoxia and hypercapnia, of 51% [31]. Accordingly, we have used idealized and patient-specific physical models to investigate howto best prevent hypoxia, hypercapnia, and lung trauma.
The Relationship Between Inlet Pressure, Flow Rate, and Mean Distal Airway Pressure
Clinical care providers currently use pressure at the ventilation needle as a reference for


59
jet ventilation. Establishing how pressure at the needle relates to distal airway pressure may help providers avoid VILI. The most fundamental parameter for invasive mechanical ventilation is the tidal volume which, as discussed above, is a function of Ppiat. There is no clear correlate in clinical jet ventilation settings so we have established the relationship between inlet pressure (Pin) and distal airway pressure (PDistai). We found a linear relationship between inlet pressure and distal airway pressure (Figure 21 and Figure 22). We established Equation 3 and Equation 4 for the ventilation models and found that the idealized airway model and CT airway model had nearly the same slope but different intercepts. These intercepts were extrapolated from the linear best fits that were based on Pin > 700 cmFLO for the CT airway model and Pin > 250 cmFhO for the idealized airway model. The relationship between inlet and distal airway pressures has a nonlinear section at very low inlet pressures. However, this nonlinearity is well below the pressures seen in clinical use.
In the low-pressure, nonlinear regime almost none of the pressure makes it to the distal airways until a critical pressure is reached. At that critical pressure, the linear section starts and a consistent portion of the inlet pressure makes it to the distal airways. Airway geometry affects the inlet pressure needed to leave the nonlinear section and start providing significant pressures in the distal airways. For the CT airway model the nonlinear section goes from 0 to 8 psi and in this section very little pressure makes it to the fifth generation airway. For the idealized airway model, the nonlinear section goes from 0 to 3 psi. The slope of the linear section was consistent between the blocks despite the numerous difference in geometry.
However, there is no inherent relationship between inlet pressure and Pdistai; this


60
correlation is contingent upon the physical configuration of the mechanical ventilation apparatus. For example, changes in needle gauge will affect flow rate and the flow rate is directly reflected in Postal. Figure 29 shows the pressure in the distal airways with the same parameters as Figure 27a but the 14-gauge needle was switched for an 18-gauge needle. This resulted in a 50.4% decrease in flow rate. The same inlet pressure and position showed a 65.7% decrease in distal airway pressure because of the reduction in flow rate. We assessed the system at three different depths, one, four, and seven centimeters into the trachea and found a 2nd degree polynomial relationship between the flow rate and distal pressure. The equations describing the relationship are seen in Figure 18.
Inlet pressure can only be used as a reference for ventilation when it can be reliably related to flow rate. This means that major and minor fluid dynamic losses must be controlled for. These sources include needle parameters such as gauge, length, and any bends or curves in the needle; supply hose parameters such as diameter, coupling type, compliance, and length past the pressure regulator; and other sources of loss in the system including valve size, shape, and locations. These sources of loss must all be kept constant between tests so that the same pressure will provide the same flow rate. During a surgical procedure, different providers have different preferences in laryngoscopes, needle types, needle mounting positions, and overall setup. It is highly unlikely that the variables responsible for dynamic losses would be consistent from procedure to procedure. Based on these findings, we assert that the pressure of the
supply line should not be used as the reference for jet ventilation.


61
Jet Flowmeter Development and Characterization
In order to address the fact that Pin and PDistai are not correlated across different ventilation hardware configurations we have developed a flowmeter to help clinicians determine appropriate settings for their patients. Inline flow meters currently on the market can accurately determine the flow rate coming out of the ventilation needle. However, since LFJV is an open system the total flow rate delivered during jet ventilation must include both the flow coming out of the needle and additional outside air that is entrained into the flow. In some cases, the entrained flow is a small portion of the total flow. In other cases, the entrained flow makes up the vast majority of the total flow [27]. The ratio of flow from the needle to entrained flow is dependent on needle gauge, jet exit velocity, and other jet flow properties such as turbulence, and interactions with other surfaces, such as a laryngoscope.
Our device allows for accurate readings at a range of flow rates and includes air entrainment. We created a characterization curve (Figure 14) for the meter and found that it has an error of less than 6% across the full range of flow rates. Above roughly 0.75 L/s the meter had an error that rarely exceeded 1%. For our experiment, which typically used 1.2 L/s to 1.4 L/s, we had a highly accurate device. The curve showed a non-linear trend in the data points and the data was tightly clustered along this curve. By creating an adjustment curve for calibrating the meter, we estimate that the error could be reduced to less than 1% across the full range of flow rates.
This device would not be able to replace the experience of the anesthesiologists in
determining the proper flow setting since this parameter varies with patient physiology. For


62
example, a pediatric ventilation case would need significantly less flow than an adult case. Many pathologic conditions would require adjustment of flow rates and even the use of different laryngoscopes would change the required flow settings. Instead, this device is meant to be a tool to grant anesthesiologists the ability to provide repeatable flow settings. With the current standard of care, even if a clinician knew that one laryngoscope provided higher distal airway pressures, trying to accurately reduce the flow by feel alone is not feasible. The ability to produce consistent flow rates from the needle and knowledge specific to the scope being used during a procedure could be extremely valuable to an experienced anesthesiologist.
Jet Needle Position and Distal Airway Pressure Heterogeneity
The radial and axial position of the jet needle in the trachea are key determinants of distal airway pressure. The relationships between depth, distance from the wall, and distal airway pressure is of major importance in emergency jet ventilation where a laryngoscope or subglottiscope is not typically used. For the sake of real time measurement and clinical relevance, rather than normalizing to flow rate, we normalized to inlet pressure. Pressure is currently the clinical norm for setting up jet ventilation system along with feeling the jet from the needle. By using the same needle and ventilation setup every time, we knew that flow rate was a simple function of pressure.
In the CT airway model, as depth into the trachea increased there was generally a counterintuitive decrease in distal airway pressure and a decrease in ventilation homogeneity. In the idealized airway model, there were no physiologically relevant changes in mean distal airway pressure with increased jet needle depth. However, the degree of heterogeneity


63
increased with jet needle depth. We looked at homogeneity between the different airways in the lung at during the same ventilation burst and how changing the position and pressure of the needle would affect the pressure at a given location in the lung.
We found that surface effects along the wall of the trachea caused a negative correlation between the distance from the tracheal wall and homogeneity of ventilation. At shallow depths into the trachea, a fully developed flow regime was able to form regardless of distance from the wall. As a result, a deeper needle placement made the model more sensitive to slight variations in needle position. A centered needle in the proximal trachea provided homogeneous ventilation more easily than a needle with a distal placement.
This is clinically relevant for determining needle placement for optimal homogeneity which in turn will improve gas exchange and reduce the risk of volutrauma and other forms of ventilator induced lung injury. In a clinical setting, the ventilation needle should ideally be centered in the trachea and as far from the carina as possible. Due the physical and visual obstruction that would be caused by having the needle centered, this is not possible during surgery. It must be placed against one of the walls. Looking at Figure 31 and Figure 33, we can see that at depth 1 the placement of the needle against the different walls makes no more than a 15% difference in the spatial mean airway pressure while homogeneity is consistent between positions.
Effects of a Laryngoscope and Subglottiscope on Distal Airway Pressure
Scopes are typically used during LFJV. They can significantly change the flow rate and
jet dynamics resulting in altered distal airway pressures. A trend of increasing pressure with


64
proximal placement of the needle seen in Figure 34 and Figure 35 continued with the addition of the subglottiscope and laryngoscope. The extended entrance length from the laryngoscope contributed to homogenous ventilation by allowing a more fully developed flow regime. Also in Figure 34 and Figure 35 we saw a 20% increase in distal airway pressure with the subglottiscope and at least a 65% increase in distal airway pressure with the Ossoff-Pilling Laryngoscope as compared to the centered needle at depth one, shown in Figure 27.
Two depths, one centimeter and three centimeters, past the laryngeal folds were tested for the subglottiscope and there was no difference in distal airway pressure or homogeneity. This means that accidentally pushing the subglottiscope a few centimeters farther than normal past the laryngeal folds during surgery would not make a substantial difference in ventilation.
A surgeon familiar with the ventilation technique set up the Ossoff-Pilling laryngoscope in a typical fashion and there was just over 10 degrees of movement in the needle within the laryngoscope. The needle was mounted at the bottom of the right side of the rectangular opening to the laryngoscope. The screw mount used to hold the needle in place prevents the needle from sliding, but allows for pivoting easily. The weight of the ventilation hose was enough to pull down on the end of the needle until the proximal end of the needle hit the laryngoscope, resulting in an upward tilt of 10 degrees. By lifting the supply hose, the needle effortlessly tilted until the tip of the needle hit the laryngoscope resulting in a level needle placement. The weight of the supply hose was the primary resistance to needle movement.
Those 10 degrees made a remarkable difference in distal airway pressure. There was an 82% increase in distal airway pressure by moving the needle across a 7-degree arc from 10


65
degrees to 3 degrees. An 82% increase in pressure could be the difference between underventilation and barotrauma. An inadvertent 7-degree needle movement could happen during a surgical procedure when tools are placed in the laryngoscope, when the evacuator is adjusted, or if the surgeon is simply unaware of the difference in pressure and bumps the needle. This means that for many jet ventilation cases the distal airway pressure will occasionally double at random intervals throughout the surgery unbeknownst to the surgeon or anesthesiologist. Unlike the laryngoscope, the subglottiscope had a secure mount for the ventilation needle and it was not possible to adjust the needle position. This promotes consistent distal airway pressure.
Conclusions and Suggestions for Clinical Practice
The current standard for low frequency jet ventilation has done many things well, but could be greatly improved with a few small changes. A quantitative measurement of total entrained flow would help anesthesiologists provide consistent, repeatable flow rates. The correlation between needle flow rate and total delivered flow rate for each laryngoscope should be established, and an equation or table should be provided with scopes to help anesthesiologists provide an ideal distal airway pressure. Future experiments should look at trends in different laryngoscopes and the appropriate flow rate for each one. Stable needle mounts should be built into the laryngoscope so the needle cannot be easily bumped, resulting in pressure spikes and drops. Improvements to laryngoscope design and flow measurement could provide more easily controlled, consistent distal airway pressures, to improve gas
exchange and reduce the risk of trauma to the lungs.


66
Future Investigations
The models designed for this experiment could be used for many follow up experiments. Characterization of additional laryngoscopes and subglottiscopes could show how consistent distal airways are within the same type of scope. Perhaps additional types of endoscopic tools such as bronchoscopes could be tested. It would also be interesting to run the same tests on additional models. Having several healthy adult models could lend additional credibility to the study and show the variety or consistency between people. Pathologic or pediatric models would also be worth investigating. The differences in equations, regression lines and data could be insightful. It might be worth adding balloons or a compliant outer section to the model to see how that effects the fluid dynamics of the block.
A particle image velocimetry experiment could elucidate what is actually happening in the airways during ventilation, especially at the carina. Perhaps this would explain why there was a counterintuitive decrease in the distal airway pressure on the same side as the needle placement. An animal model may ultimately be the most realistic form of this experiment. Placing transducers into the distal airways and measuring the actual pressures during
ventilation could be insightful.


67
REFERENCES
1. Tucker, W.D. and B. Burns, Anatomy, Thorax, Heart Pulmonary Arteries, in StatPearls. 2019.
2. Donley, E.R. and J.W. Loyd, Anatomy, Thorax, Wall Movements. 2019.
3. Patwa, A. and A. Shah, Anatomy and physiology of respiratory system relevant to anaesthesia. Indian journal of anaesthesia,
2015. 59(9): p.533-541.
4. Verschakelen, J.A. and W. De Wever, Basic Anatomy and CT of the Normal Lung, in Computed Tomography of the Lung: A Pattern Approach, J.A. Verschakelen and W. De Wever, Editors. 2018. p. 3-19.
5. Government, T.U.S., Lung Diagram. 2006.
6. Macklem, P.T., The Mechanics of Breathing. American Journal of Respiratory and Critical Care Medicine, 1998. 157(4): p. S88-S94.
7. Saraswat, V., Effects of anaesthesia technigues and drugs on pulmonary function. Indian journal of anaesthesia, 2015. 59(9): p. 557-564.
8. Knowlson, G.T.G. and H.F.M. Bassett, THE PRESSURES EXERTED ON THE TRACHEA BY ENDOTRACHEAL INFLATABLE CUFFS. British Journal of Anaesthesia, 1970. 42(10): p. 834-837.
9. Smallwood, C.D. and B.K. Walsh, Noninvasive Monitoring of Oxygen and Ventilation. Respiratory Care, 2017. 62(6): p. 751.
10. Broche, L., et al., Dynamic Mechanical Interactions Between Neighboring Airspaces Determine Cyclic Opening and Closure in Injured Lung. Critical Care Medicine, 2017. 45(4): p. 687-694.
11. Motta-Ribeiro, G., et al., Spatial Heterogeneity of Lung Strain and Aeration and Regional Inflammation During Early Lung Injury Assessed with PET/CT. Academic Radiology, 2019. 26(3): p. 313-325.


68
12. Michiels, C., Physiological and pathological responses to hypoxia. The American journal of pathology, 2004. 164(6): p. 1875-1882.
13. S, P. and M. SH, Physiology, Carbon Dioxide Retention. StatPearls [Internet], (2019 Jan-).
14. Heinz, U.E. and J.D. Rollnik, Outcome and prognosis of hypoxic brain damage patients undergoing neurological early rehabilitation. BMC research notes, 2015. 8: p. 243-243.
15. Gattinoni, L., et al., Relationships between lung computed tomographic density, gas exchange, and PEEP in acute respiratory failure. Anesthesiology, 1988. 69(6): p. 824-32.
16. Bilek, A.M., K.C. Dee, and D.P. Gaver III, Mechanisms of surface-tension-induced epithelial cell damage in a model of pulmonary airway reopening. J. Appl. Physiol., 2003. 94: p. 770-783.
17. Kay, S.S., et al., Pressure gradient, not exposure duration, determines the extent of epithelial cell damage in a model of pulmonary airway reopening. J. Appl. Physiol., 2004. 97: p. 269-276.
18. Mead, J., T. Takishima, and D. Leith, Stress distribution in lungs: a model of pulmonary elasticity. Journal of applied physiology, 1970. 28(5): p. 596-608.
19. Makiyama, A.M., et al., Stress concentration around an atelectatic region: A finite element model. Respir Physiol Neurobiol, 2014.
201: p. 101-10.
20. Tremblay, L.N. and A.S. Slutsky, Ventilator-induced lung injury: from the bench to the bedside. Intensive Care Med, 2006. 32(1): p. 24-33.
21. Hamlington, K.L., et al., Predicting ventilator-induced lung injury using a lung injury cost function. J Appl Physiol (1985), 2016.
121(1): p. 106-14.
22. Henderson, W.R., et al., Fifty Years of Research in ARDS.Respiratory Mechanics in Acute Respiratory Distress Syndrome. American


69
Journal of Respiratory and Critical Care Medicine, 2017. 196(7): p. 822-833.
23. Baumgardner, J.E., et al., Effects of Respiratory Rate, Plateau Pressure, and Positive End-Expiratory Pressure on Pa02 Oscillations after Saline Lavage. American Journal of Respiratory and Critical Care Medicine, 2002. 166(12): p. 1556-1562.
24. Smith, B.J. and J.H.T. Bates, Assessing the Progression of Ventilator-Induced Lung Injury in Mice. IEEE Trans. Biomed. Eng., 2013.
60(12): p. 3449-3457.
25. Otto, C.M., et al., Spatial and temporal heterogeneity of ventilator-associated lung injury after surfactant depletion. Journal of applied physiology (Bethesda, Md. : 1985), 2008. 104(5): p. 1485-1494.
26. Turner, J.S., Turbulent entrainment: the development of the entrainment assumption, and its application to geophysical flows. Journal of Fluid Mechanics, 1986. 173: p. 431-471.
27. Van Der Spek, A.F., P.M. Spargo, and M.L. Norton, The physics of lasers and implications for their use during airway surgery. Br J Anaesth, 1988. 60(6): p. 709-29.
28. Evans, E., P. Biro, and N. Bedforth, Jet ventilation. BJA Education, 2007. 7(1): p. 2-5.
29. Baraka, A.S., et al., Low freguency jet ventilation for stent insertion in a patient with tracheal stenosis. Canadian Journal of Anesthesia, 2001.48(7): p. 701-704.
30. Pathak, V., et al., Ventilation and Anesthetic Approaches for Rigid Bronchoscopy. Annals of the American Thoracic Society, 2014. 11(4): p. 628-634.
31. Duggan, L.V., et al., Transtracheal jet ventilation in the 'can't intubate can't oxygenate' emergency: a systematic review. Br J Anaesth, 2016. 117 Suppl 1: p. i28-i38.
32. Care, U.o.I.H. Jet Ventilation Anesthesia - Transoral for Laryngeal Surgery. Iowa Head and Neck Protocols 2019 6/18/2019];


70
Available from: https://medicine.uiowa.edu/iowaprotocols/iet-ventilation-anesthesia-transoral-laryngeal-surgery.
33. UCSF, D.o.A.P.C. Manual Jet Ventilation. 2013.
34. Reed, J.P., et alv Studies with transtracheal artificial respiration. Anesthesiology, 1954. 15(1): p. 28-41.
35. Chhangani, S.V., CHAPTER 30 - Independent Lung Ventilation and Bronchopleural Fistula, in Mechanical Ventilation, P.J. Papadakos, B. Lachmann, and L. Visser-lsles, Editors. 2008. p. 337-354.
36. Putz, L., A. Mayne, and A.-S. Dincq, Jet ventilation during rigid bronchoscopy in adults : a focused review.". Vol. 2016. 2016.
37. Klain, M. and R.B. Smith, High freguency percutaneous transtracheal jet ventilation. Crit Care Med, 1977. 5(6): p. 280-7.
38. Pawlowski, J., Anesthetic considerations for interventional pulmonary procedures. Curr Opin Anaesthesiol, 2013. 26(1): p. 6-12.
39. Bouroche, G., et al., Rescue transtracheal jet ventilation during difficult intubation in patients with upper airway cancer. Anaesth Crit Care Pain Med, 2018. 37(6): p. 539-544.
40. Noonan, M., et al., Intraoperative High-Freguency Jet Ventilation Is Eguivalent to Conventional Ventilation During Patent Ductus Arteriosus Ligation. World J Pediatr Congenit Heart Surg, 2017. 8(5): p. 570-574.
41. Abad, H.L., M. Ajalloueyan, and A.R. Jalali, Impact of body mass index (BMI) on ventilation during low-freguency jet ventilation. Otolaryngol Head Neck Surg, 2007. 136(3): p. 477-80.
42. Altun, D., et al., Surgical Excision of Postintubation Granuloma Under Jet Ventilation. Turkish journal of anaesthesiology and reanimation, 2014. 42(4): p. 220-222.
43. Philips, R., B. deSilva, and L. Matrka, Jet ventilation in obese patients undergoing airway surgery for subglottic and tracheal stenosis. Laryngoscope, 2018. 128(8): p. 1887-1892.


71
44. Goudra, B.G., et al., Anesthesia for Advanced Bronchoscopic Procedures: State-of-the-Art Review. Lung, 2015. 193(4): p. 453-465.
45. Bourgain, J.L., et al., [Guide for the use of jet-ventilation during ENT and oral surgery]. Ann Fr Anesth Reanim, 2010. 29(10): p. 720-7.
46. Florens, M., B. Sapoval, and M. Filoche, An anatomical and functional model of the human tracheobronchial tree. J Appl Physiol (1985), 2011. 110(3): p. 756-63.
47. Horsfield, K., et al., Models of the human bronchial tree. J Appl Physiol, 1971. 31(2): p. 207-17.
48. Weibel, E.R., Morphometry of the Human Lung, ed. E.R. Weibel. 1963, 1-151.
49. Fernandez-Bustamante, A., et al., High-freguency jet ventilation in interventional bronchoscopy: factors with predictive value on high-freguency jet ventilation complications. J Clin Anesth, 2006. 18(5): p. 349-56.
50. Hautmann, H., et al., High freguency jet ventilation in interventional fiberoptic bronchoscopy. Anesth Analg, 2000. 90(6): p. 1436-40.
51. Conacher, I.D., Anaesthesia and tracheobronchial stenting for central airway obstruction in adults. Br J Anaesth, 2003. 90(3): p. 367-74.


72
APPENDIX
Depth into Lung Cast (cm)
Figure 36: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the dorsal wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Dorsal 1 is closest to the center and dorsal 3 is closest to
the tracheal wall.


73
Figure 37: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the ventral wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Ventral 1 is closest to the center and ventral 3 is closest to
the tracheal wall.


74
Figure 38: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the
tracheal wall.


75
Figure 39: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the right wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Right 1 is closest to the center and right 3 is closest to the
tracheal wall.


76
Figure 40: Homogeneity in distal airway pressure in the CT airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.


77
Depth into Lung Cast (cm)
Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and Left 3 is closest to the
tracheal wall.


78
5.5
5
=3 tft cft CD
i_
Q_
4.5
_N
E
o
4
3.5
012345673 Depth into Lung Cast (cm)
Figure 41: Homogeneity in distal airway pressure in the idealized airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.
x10'3
1---------------------1--------------------1---------------------1--------------------1---------------------1--------------------T
Ventral
Dorsal
Left
Right


79
Figure 42: Homogeneity grid for ventilation in the idealized airway block with the needle placed
in the far right position in the trachea at depth 1.


80
Pressure Homogeneity for Depth 4, Right 3 xio-3^
Figure 43: Homogeneity grid for ventilation in the idealized airway block with the needle
placed in the far right position in the trachea at depth 4. Bimodal ventilation is seen here.


81
Figure 44: Ventilation in the idealized airway block with the needle placed in the far right
position in the trachea at depth 7. Bimodal ventilation is seen here.


82
DECLARATION OF ORIGINAL WORK
This page is to affirm that all work in this Master’s Thesis is my original work. Further, I confirm that all writing is my own writing. Work from others has been cited appropriately.
Joshua Pertile 7/19/19
Student Name Date


Full Text

PAGE 1

PARAMETERS GOVERNING AIRWAY PRESSURE HOMOGENEITY DURING LOW FREQUENCY JET VENTILAITON By JOSHUA DANIEL PERTILE B.S., University of Wisconsin, 2017 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science in Bioengineering Bioengineering Program 2019

PAGE 2

ii © 2019 JOSHUA DANIEL PERTILE ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Master of Science Bioengineering degree by Joshua Daniel Pertile has been approved for the Bioengineering Program by Bradford Smith, Chair Daniel Fink Kendall Hunter Date: July 27 th , 2019

PAGE 4

iv Pertile, Joshua Daniel (M.S., Bioengineering) Parameters Governing Airway Pressure Homogeneity During Low Frequency Jet Ventilation Thesis directed by Ass istant Professor Bradford Smith ABSTRACT Low frequency jet ventilation (LFJV) is a widely used fo rm of ventilation that currently has no defined safe pressure s or flow rate s because the methodology for safe and effective ventilation has been insufficiently researched. This study looked at pressure in the distal airways during low frequency jet ventilation and how different parameters contributed to an even pressure distribution across the lung ( pressure homogeneity ) . Two physical models of airway trees were fabricated : one bas ed on a CT scan of a healthy 17 year old mal e and second that was a symmetric, idealized representation of the human airway tree . We found that needle placement closer to the proximal end of the trachea and centered within the airway cross section produced the most homogenous pressures at the 5 th a irway generation. Additionally, the more distal the needle placement was in the trachea, the more sensitive the distal airway pressure was to the needle being off the tracheal centerline . The distal airway pressure is determined by total entrained flow rate and not just pressure supplied to the ventilation needle. T he relationship between flow rate and distal airway pressure is a second order polynomial with R 2 > .99. Supply pressure at t he needle is a determining factor in flow rate but in itself is not sufficient to determine distal airway pressure. A device was created to measure the total entrained flow rate and the design considerations are detailed herein .

PAGE 5

v The distal airway pressure s during ventilation with a clinical laryngoscope and subglottis cope were also measured . The depth of scope placement from one to three centimeters did not substantially affect distal airway pressure or homogeneity. The placement of the needle in the lar yngoscope substantively changed distal airway pressure. Even with consistent needle depth , a needle angle change of a few degrees nearly double d the distal airway pressure s . We conclude that l aryngoscope designs should be updated to securely hold the ventilation needle to provide consistent airway pressures. Furthermore, the jet flow rate should be used to set inlet pressure prior to ventilation so that distal airway pressures are consistent across different ventilation hardware configurations . The for m and content of this abstract are approved. I recommend its publication. Approved: Bradford Smith

PAGE 6

vi To all those who instilled in me a sense of curiosity , especially my parents

PAGE 7

vii Acknowledgements I would like to acknowledge Bradford Smith for scientific and material support and mentoring me though the thesis process , Daniel Fink for funding and advice , Jennifer Wagner and Emily DeBoer for providing the CT airway model and assisting in the fabrication of the idealized airway model, and Michelle Mellenthin for countless hours of help with programming and circuit design and supplying hundreds of cups of high quality coffee.

PAGE 8

viii Table of Contents CHAPTER I. INTRODUCTION ................................ ................................ ................................ ................... 1 II. BACKGROUND ................................ ................................ ................................ ..................... 7 III. MATERIALS AND METHODS ................................ ................................ .............................. 12 Model Airway Pressure Measurements ................................ ................................ ........ 12 Jet Needle Positioning System ................................ ................................ ................... 15 Signal and Data Processing ................................ ................................ ........................ 17 Airway Pressure Measurement Procedure ................................ ............................... 21 Flow Measurement ................................ ................................ ................................ ....... 22 Additional Consi derations ................................ ................................ ......................... 26 IV. RESULTS ................................ ................................ ................................ ............................ 29 Flowmeter Characterization ................................ ................................ .......................... 29 Airway Pressure Measurements ................................ ................................ ................... 30 Data Normalization ................................ ................................ ................................ ....... 33 Airway Pressure Relationships ................................ ................................ ...................... 35 Airway Pressure Distribution ................................ ................................ ........................ 40 Lary ngoscopes and Distal Airway Pressure ................................ ................................ ... 53

PAGE 9

ix V. DISCUSSION ................................ ................................ ................................ ...................... 57 The Relationship Between Inlet Pressure, Flow Rate, and Mean Distal Airway Pressure ................................ ................................ ................................ ................................ ................... 58 Jet Flowmeter Development and Characterization ................................ ...................... 61 Jet Needle Position and Distal A irway Pressure Heterogeneity ................................ ... 62 Effects of a Laryngoscope and S ubglottiscope on Distal Airway Pressure ................... 63 Conclusions and Suggestions for Clinical Practice ................................ ........................ 65 Future Investigations ................................ ................................ ................................ ..... 66 REFERENCES ................................ ................................ ................................ ................................ .. 67 APPENDIX ................................ ................................ ................................ ................................ ...... 72

PAGE 10

x List of Figures Figure 1: The anatomy of the human lung [5] ................................ ................................ ............... 2 Figure 2: Air entrainment in a laryngoscope ................................ ................................ ................. 5 Figure 3: A standard jet ventilation setup. The red arrow is pointing to the pressure regulator. 8 Figure 4 : Idealized airway geometry based on the morphometry models from Florens, Horsfield, and Weibel [46 48]. ................................ ................................ .......................... 13 Figure 5: Model of the human airway tree, derived from the CT scan of a 17 year old healthy human subject. ................................ ................................ ................................ ................. 14 Figure 6: Photo of the experimental setup with the CT airway block. The regulator is indicated with a red arrow. ................................ ................................ ................................ ............... 16 Figure 7: Wiring diagram for the data acquisition system from the pressure transducers to the computer. ................................ ................................ ................................ .......................... 19 Figure 8: A top down view of the trachea with the thirteen ventilation positions shown: The center, and three steps in each direction. ................................ ................................ ........ 22 Figure 9: Venturi flow meter design. The locations of A 1 , A 2 , P 1 , and P 2 (Eq. 1) are labeled in the cutaway (A). ................................ ................................ ................................ ...................... 23 Figure 10: Jet needle mounting system ................................ ................................ ........................ 24 Figure 11: The curve of the flow rate vs the measured voltage. This provides an accurate measurement from a flow rate of 0.25 L/s to approximately 3.25 L/s. ........................... 26 Figure 12: Twist on Mounting Bracket ................................ ................................ .......................... 27 Figure 13: Functioning prototype of the air flow entrainment measuring device, with housing

PAGE 11

xi removed ................................ ................................ ................................ ............................ 28 Figure 14: Volume determined as the integral of the flow measured using the venturi flowmeter (blue circles). The target value for each measurement is 3 L (red line) that is delivered using a calibrated syringe. ................................ ................................ ................................ 30 Figure 15: The demultiplexed, unfiltered, uncalibrated voltage readings from each sensor during a ventilation test run on the CT airway block. The needle was centered and moved through all 7 depths. D1 D7 denote the depth into the trachea from 1cm to 7cm. ....... 31 Figure 16: Distal airway pressure measuremen ts in the CT airway block with the jet needle 6mm from the left wall of the trachea (first step to the left) showing the filtered pressure for all 16 distal airway sensors (a). The homogeneity graph (b) shows the mean pressure (heavy black line) and the sh aded area represents the range from the highest to lowest mean value for an individual sensor at that depth. Note that (b) is normalized by the inlet pressure. D1 D7 indicate depth into the trachea from 1cm to 7cm. ...................... 33 Figure 17: Transient in inlet pressure over the three second ventilation burst provided during the experiment. This is measured by a pressure transducer placed immediate ly before the ventilation needle. ................................ ................................ ................................ ...... 34 Figure 18: Relationship between distal airway pressure and total flow rate including entrained a ir. This is in the center position of the idealized airway block at 1 (a), 4 (b), and 7 cm (c) into the trachea. ................................ ................................ ................................ ............... 36 Figure 19: Relat ionship between distal airway pressure and total flow rate shown at 1, 4, and 7 cm into the trachea, with the needle centered in the idealized airway block. ................ 37 Figure 20: Distal airway pressure vs inlet pressure for the center needle placement in the

PAGE 12

xii idealized airway model at depths of 1 (a), 4 (b), and 7 cm (c). ................................ ....... 38 Figure 21: The linear regression of inlet pressure vs distal airway pressure for the CT airway model with the needle centered in the trachea at depth 1. ................................ ............ 39 Figure 22: The linear regression of inlet vs distal pressure for the idealized airway model with the needle centered in the trachea at depth 1. ................................ ............................... 40 Figure 23: The numbered positions of the sensors on the idealized airway block ...................... 41 Figure 24: Sensor represented on the homogeneity grid. ................................ ........................... 42 Figure 25: Example of airway bifurcations to reach sensor 9. Each darker shade represents a branch farther into the block. ................................ ................................ ........................... 43 Figure 26: Comparison of homogeneity in the distal airways for the idealized airway model at depths 1 (a), 4 (b), and 7 cm (c). ................................ ................................ ....................... 44 Figure 27: The homogeneity grids at depths 1 (a), 4 (b), and 7 cm (c) with the jet needle centered in the trachea in the CT airway block. ................................ .............................. 45 Figure 28: Trimodal ventilation demonstrated in the CT airway block with the needle placed in the far right position in the trachea at depths of 1 (a), 4 (b) , and 7 cm (c). ..................... 47 Figure 29: The homogeneity grid at depth 1 with an 18 gauge needle in the center of the trachea for the CT airway tree. ................................ ................................ ......................... 47 Figure 30: Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left side of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shade d area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps

PAGE 13

xiii toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wal l. ................................ ........................ 49 Figure 31: Comparison of needle placements in the idealized airway block for the center and the placements closest to the trac heal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean v alue for an individual sensor at that depth. ............................. 50 Figure 32: Homogeneity in the distal airways of the CT airway block as a function of distance from the left side of the trachea. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wall. ................................ ................................ ................................ .............. 51 Figure 33: Comparison of needle placements in the CT airway block for the center and the placements closest to the tracheal wall in each direction. The dark line in the ce nter for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. ............................. 52 Figure 34: Pressure homogeneity grids for the subglottiscope with the orifice 1 and 3 centimeters past the laryngeal folds. This is in the CT airway block. ............................... 54 Figure 35: The homogeneity grids for the Ossoff Pilling Laryngoscope ventilation. The needle is

PAGE 14

xiv mounted at the bottom of the right wall and tilted up at angles of 10, 5, 3, and 0 degrees. This is the CT airway model. ................................ ................................ .............. 55 Figure 36: Homogeneit y in distal airway pressure in the CT airway block as a function of depth as the distance from the dorsal wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three s econds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Dorsal 1 is closest to the center and dorsal 3 is closest to the tracheal wall. ................................ .................... 72 Figure 37: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the ventral wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Ventral 1 is closest to the center and ventral 3 is closest to the tracheal wal l. ................................ .............. 73 Figure 38: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shade d area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center

PAGE 15

xv and left 3 is closest to the tracheal wal l. ................................ ................................ .......... 74 Figure 39: Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the right wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mea n value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Right 1 is closest to the center and right 3 is closest to the tracheal wall. ................................ ............................. 75 Figure 40: Homogeneity in distal airway pressure in the CT airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the cente r for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. ............................. 76 Figure 41: Homogeneity in distal airway pressure in the idealized airway block as a function of depth in the position closest to tracheal w all in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value fo r an individual sensor at that depth. ............. 78 Figure 42: Homogeneity grid for ventilation in the idealized airway block with the needle pl aced in the far right position in the trachea at depth 1. ................................ ........................... 79 Figure 43: Homogeneity grid for ventilation in the idealized airway block with the needle placed in the far right position in the trachea at depth 4. Bimodal ventilation is seen

PAGE 16

xvi here. ................................ ................................ ................................ ................................ .. 80 Figure 44: Ventilation in the idealized airway block with the needle placed in the far right position in the trachea at depth 7. Bimodal ventilation is seen here. ............................ 81

PAGE 17

1 CHAPTER I I NTRODUCTION The lung is where gas exchange occurs in the body. In order to maximize gas exchange, the anatomy of the lung maximizes the surface area of the air blood interface in a region of the lung known as the parenchyma where gas exchange occurs . The parenchyma is made up of fragile , microscopic sacs that fill with air, called alveoli [1, 2] and the walls of the alveoli contain a dense network of capillaries . Miniscule airways known as the alveolar ducts supply air to the alveoli, and the respiratory bronchioles supply air to the alveolar ducts. They receive air from the two bronchi, which bifurcate from the trachea the primary air supply lumen for the lungs . At the top of t he trachea, there is a muscular section referred to as the larynx , that includes two folds called the laryngeal folds , which protect the airway, and provide phonation, or speech [3, 4] . Figure 1 shows a diagram of the anatomy of the human lung.

PAGE 18

2 Figure 1 : The anatomy of the human lung [5] Respiration is driven by the diaphragm, a dome shaped muscle located at the base of the lungs , and the intercostal muscles, which run between the ribs . When the diaphragm contracts, is pulls d own on the lungs and the intercostal muscles expand the chest wall to reduce the plural pressure [2, 6] . This negative pressure pulls the lungs outward, lowering alveolar pressure and drawing in ambient air to fill the lungs. Pressure in airways during normal, relaxed breathing will not greatly exceed atmospheric pressure, and due to the mechanics of respi ration there is no way for natural breathing to fill the lungs beyond capacity. During surgery, anesthesia is used to prevent pain and discomfort in the patient. Drugs used in a nesthesia include analgesics, sedatives, and paralytics, which can all reduc e or prevent

PAGE 19

3 normal respiration , so gas exchange must be artificially provided via mechanical ventilation [7] . The airway is typically maintained through endotracheal intubation, which provides a path for exchange of oxygen and carbon dioxide and introduction of anesthetic gasses. The endotracheal (ET) tube is sealed in the airw ay with an inflatable cuff. The ET cuff is a balloon like device surrounding the ET tube that is inflated to expand in the trachea, pushing against the tracheal wall to create a seal and provide a closed ventilation system [8] . Airway p ressure can be maintained at appropriate levels and the mixture of inspired gas can be set as necessary for the procedure. Exhaled carbon dioxide levels may be monitor ed as well to ensure that proper gas exchang e is occurring [9] . Modern clinical ventilators use positive pressure to force air into th e lungs and , as such, it is possible for the lung s to be filled beyond their natural capacity. In addition, h eterogeneous ventilation can , and does , regularly occur during mechanical ventilation [10, 11] . This is when an airway receives excessive or insufficient ventilatio n relative to the surrounding airways . In the case of excessive ventilation, the airway or alveoli experience large volume s and pressure s that can distend and damag e the tissue . Insufficient ventilation results in reduced gas ex change between the blood a nd air in the alveolar sacs, resulting in hypoxemia and hypercapnia . Hypoxemia and hypercapnia can cause complications ranging from temporary organ dysfunctions to death [12 14] . In addition, low airway pressures may result in the collapse, or derecruitment, of small airways and alveoli [15] . I njury caused by ventilation is referred to as ventilator induced lung injury, or VILI. Atelectatic regions in the lungs will cyclically recruit and derecruit resulting in atelectrauma.

PAGE 20

4 Atele ctrauma is caused by a repeated fluid mechanical stress on the alveolar walls that cause direct injury to the epithelium of alveoli and small airways [16, 17] . Injury caused by o verdistention of the alveoli is termed volutrauma . Because alveoli share septal walls with their adjacent neighbors there is a mechani cal interdependence between alveoli. This may potentiate volutrauma during heterogeneous ventilation where inflated alveoli are adjacent to poorly aerated alveoli [18, 19] . The goals of preventing volutrauma and atelectrauma frequently require conflict ing ventilation settings . As such, v entilation is a balancing act of maintaining high enough pressure and tidal volume to maintain gas exchange, and high enough PEEP to prevent atelectrauma, while keeping pressure low enough to avoid volutrauma [20 23] . Furthermore, VILI forms its own positive feedback loop. If an airway is collapsed or flooded, it will promote further inj ury . As VILI progresses, recruitment and decruitment occur more frequently at higher pressures promoting further atelectrauma [24] . Flooded or collapsed sections of the lung promote heterogeneous ventilation in the dependent airways, further reduc ing the homogeneity of ventilation , and pro moting more damag e [25] . As damage accrues in the lung, gas exchange is reduced and more aggressive ventilation is required to prevent hypercapnia and hypoxia. Endotracheal intubation cannot be used in upper airway procedures such as subglottic and tracheal stenosis surgeries , rigid bronchoscopies, airway granuloma removals, and many other procedures where airway access is necessary. Intubation block s access to the airway and an adequate sea l from the ET cuff may not be possible if the integrity of the tracheal wall is compromi sed during the procedure . An open ventilation system must be used, typically in the

PAGE 21

5 form of jet ventilation. During LFJV , a needle connected to a high pressure supply line is typically placed in an endoscopic device, such as a laryngoscope or a subglotti scope. L aryngoscope s and subglottiscope s are tools used to hold the airway open and allow for surgical and diagnostic tools to be placed in the airway. The subglottiscope allows for deeper placement into the airway than the laryngoscope if needed or prefe rred by the surgeon . A ll forms of j et ventilation deliver a jet of high pressure air down the trachea and ambient air is entrained in the flow. Figure 2 : Air entrainment in a laryngoscope Entrainment is a hydrodynamic property of fluids to carry, or entrain, surrounding fluids into their flow [26] . In the case of jet ventilation, t he high velocity, turbulent jet creates a low pressure area, similar to the Bernoulli effect, and ambient air is pulled into the path of the jet [27, 28] . This is shown in Figure 2 . This means that distal airway pressure is difficult to control because the volume of air delivered is a function of the supplied air , ambient air that is pulled down the tr achea by the jet , and air that flows back out the trachea and past the incoming jet . The total flow from a laryngoscope including entrainment has been measured at 20 times the

PAGE 22

6 flow produced by the jet [27] . The lack of control makes consistent ventilation difficult and may increas e the occurrence of VILI. M ethods to provide safe, homogeneous ventilation via jet ventilation have been insufficiently studied. Our long term goal is to make pressure adjustments for LFJV more of a science and less of an art. We postulate that evidence based written standard s for LFJV and a device to produce repeatable flow readings will help to optimize practices and reduce associated guesswork for anesthesiologists using the technique, improving patient outcomes.

PAGE 23

7 CHAPTER II B ACKGROUND Low frequency jet ventilation (LFJV) is a w idely used form of ventilation. It is frequently used during rigid bronchoscopies scenarios, and surgery for laryngo tracheal stenosis where intubation is not possible or would obscure the surgic al field [29 31] . In rigid bronchoscopy, LFJV is currently the most commonly used form of ventilation [30] . While statistic s on the number of LFJV cases per year are not readily available, based on the frequency of use in the procedures listed above, we estimate there are 200,000 LFJV cases in the US per year. Pressure adjustment for jet ventilation is currentl y performed empirically using a low precision pressure regulator , shown in Figure 3 . The pressure is then tested by applying air from the ventilation needle onto the a [27] .

PAGE 24

8 Figure 3 : A standard jet ventilation setup. The red arrow is pointing to the pressure regulator. While LFJV remains in widespread use, there is limited data on settings that optimiz e oxygenation and lung protection using this technique. Without continuous capnography provided by a closed anesthetic circuit, it is not possible to accurately measure end tidal CO 2 concentrations in real time and thus adequacy of ventilation is based on intermittent blood gas concentrat ion sampling [28] . There is currently no written standard for l ow frequency jet ventilation pressures and flow rates , leaving anesthesiologists to rely on personal experience and training rather than evidence based guidelines . As such, methods to set pressure var y by provider and institution. Many set by feeling for the correct pressure on their forearm while others start at a low pressure and then slowly increase pressure during ventilation until they see adequate chest rise [27, 32] . The University of California San Francisco recommends

PAGE 25

9 starting with a flat setting of 50 PSI for adult patients [33] . Applying insufficient pressure can result in hypoventilation leading to hypoxia and hypercapn ia w hich can cause organ failure, permanent brain damage, and in extreme cases , death [12 14] . Applying excess pressure can result in volutrauma while i nsufficient PEEP can cause atelectrauma [16, 17] . The lack of data based standards and hardware to allow repeatable jet settings l eads many providers to avoid jet ventilation altogether even when it may provide the optimal form of ventilation for a given procedure. Moreover, the use of qualitative jet settings makes it difficult to determine and share effect ive parameters between providers and institutions. Low frequency jet ventilation was introduced in the 1950s as a method of non intrusive ventilation during upper airway surgery [34] . High frequency jet ventilation, another form of jet ventilation, was deve loped shortly after and uses low tidal volumes at high respiratory rates of 100 600 breaths/min [35, 36] . The first measurement of jet ventilation pressures was over 20 years later, in 1977, when tracheal pressure was measured and it was determined that high frequency jet ventilation could generate small amounts of posi tive end expiatory pressure ( PEEP ) [37] . Recommended safe jet i nlet pressures range from 20 to 50 psi depending on the institution [32, 33, 36] . Note that these inlet pressures are three to four orders of magnitude above the distal airway pressure. L ike other forms of ventilation, there are complications ass ociated with LFJV including hypotension, hypertension, hypercapnia, and hypoxemia [38] . Studies of LFJV during rigid bronchoscopy found that oxygen ation can be maintained at adequate levels during ventilation while barotrauma associated complications , such as cervical emphysema and tension

PAGE 26

10 pneumothorax, remain under 1% [30] . Another study showed that j et ventilation is an effective method of rescue ventilation during o perations for throat cancers . That study established advantages in decreased surgical obstruction and increased gas exchang e compared to rescue by mask ventilation by successfully restoring proper blood oxygen levels in all 31 patients included in the study [39] . For ductus arteriosus ligation, there was no significant difference in gas exchange, or rate of complications between jet ventilation and other forms of ventilation [40] . Jet ventilation is currently the gold standard for 'can't intubate can't ventilate' situations but the rate s of barotrauma and total complications may be as high as 32% and 51% respectively, even in situation s where jet ventilation i s considered the best method [31] . T hat a literature review that noted how frequently ot her sour ces listed barotrauma as a complication and specifically included pneumothorax, p neumomediastinum , and subcutaneous emphysema as complications . Note that these barotrauma rates are much higher than reported during rigid bronchoscopy and this may be due to the challenging situation where emergency ventilation is applied. Obesity has been shown to have little to no correlation with complicatio ns during jet ventilation [41 43] . This is a somewhat surprising result given that the increased weight of the chest in obese subjects is known to reduce respiratory system compliance, and thus tidal volumes , at a given pressure. Although there have been a number of studies of LFJV , none have directly addressed what is possibly the most important question : h ow can the technique be applied in a consistently safe and effective manner? As such, a nesthesiologists still have no evidence based, quantitative guide lines for safe

PAGE 27

11 pressure or flow rate settings. Furthermore, the jet ventilation pressures are set according to the inlet pressure supplied to the needle and by feeling the jet of air coming out of the needle [44] . Since the flow rate delivered to the patient also depends on equipment downstream of the pressure regulator , we assert that this is not a suitable approach. These are major shortcoming since i mproper pressures or techniques may lead to heterogeneous ventilation, barotrauma, impaired gas exchange, and pneumothorax [45] .

PAGE 28

12 CHAPTER III MATERIALS AND METHODS We designed airway models to explore how jet pressure, flow rate, needle placement in the trachea, and laryngoscope and subglottiscope use affected distal airway pressure. These models were fitted with pressure transducers and then circuitry and software was created to record pressure data. A flow measurement device wa s developed to quantify flow rate including flow entrainment. Our experimental setup was designed to characterize the fluid dynamic properties seen in human airways relevant to LFJV. Model Airway Pressure Measurements The experiment uses two different p hysical models of the upper airway. The first model is an idealized representation of the first five generations of the human airway tree and is Airway . Th e design is unique to this experiment and the branch angle, s ize, a nd generational restriction s are derived from works by Florens, Horsfield, and Weibel et al . [46 48] . Our model is symmetric in the sagittal and coronal planes. The first generation is a 125mm long and 18mm d iameter trachea . Each of the four bifurcations has a 70 degree branch angle and t he plane of each bifurcation is rotated perpendicular to the previous bifurcation about the axi s of the superior duct. The diameter at each generation of the airway was 79% of th e size of diameter of the superior airway. Each generation, except for the trachea, maintained a diameter to length ratio of 1:3. This idealized airway model allows the di stal airway pressure relationship to be studied purely as a function of needle placement in

PAGE 29

13 the trachea. The symmetry in the block also allows for verification of results by checking that symmetrical ventilation produces symmetrical results. Figure 4 : Idealized airway geometry based on the morphometry model s from Florens, Horsfield , and Weibe l [46 48] . We also consider ed a full size fabrication based on a CT s can from a healthy 17 year old male human subject, shown in Figure 5 . This model , Airway , was originally developed for bronchoscopy training and was kindly donated by Emily DeBoer and Je n nifer Wagner of the University of Colorado . The length of the trachea was 129mm.

PAGE 30

14 Figure 5 : Model of the human airway tree, derived from the CT scan of a 1 7 year old healthy human subject. The idealized and CT airway models are each comprised of a solid block of 50 durometer silicone with a hollow airway inside. The idealized airway was designed in SolidWorks, and the CT scan was imported. Positive casts were then 3D printed out of c alcium sulfate hemihydrate with a Projet CJP 660Pro (3DSystem, USA) with and surface coated w ith paraffin wax. The print was encased in silicone and then dissolved using a heated ultrasonic bath, creating a hollow airway in the silicone block. Holes for access ports were drilled in the block s to reach the fifth airway generation in each lobe of the lung in the CT airway model, and each of the sixteen distal airways for the idealized airway model . Silicone adhesive was used to mount barbs at the

PAGE 31

15 ports to allow sensors to be easily attached and removed. This created a sealed airway that allowed for easy measurement in the distal airways. Jet Needle Positioning System In order to investigate the effects of different jet positions in the trachea , an automated system was designed and fabricated to control the position of the ven tilation needle. A n Anet A8 3D printer with a precision of 0.2mm was used as the base of the system and a fixture , shown in Figure 6 , was designed to hold the lung cast in place on the printing platform. The fixture was bolted to the platform of the 3D printer and wheels were attached to allow for smooth movement. The main printing code was deleted from the printer and G code was writ ten for movements needed for ventilation. A fixture was designed to hold a Luer Lock ventilation needle where the printer extrusion head was originally located . For this experiment, a 100 mm 14 gauge needle was used , as is typical for jet ventilati on. The mount was designed to hold the needle level and pointing straight ahead for maximum stability and repeatability . The gas delivery system wa s configured with an electric solenoid valve and a manual button in parallel to allow for electrically actu ated or manual ly triggered ventilation . The electric valve was connected to the printer control system so that ventilation experiments were fully automated. The system is shown with the CT airway block in Figure 6 .

PAGE 32

16 Figure 6 : Photo of the experimental setup with the CT airway block. The regulator is indicated with a red arrow. The system was tested for consistency and can make repeatable maneuvers . Test measurements of repeated runs in the same position had consistent distal airway pressures with a standard deviation of less than a percent difference. The blocks were highly sensitive to cha nges in position. In one case, a misalignment of the block on the ventilation platform caused the block to be rotated about the trachea by about 3 degrees. This resulted in substantial changes in distal airway pressure. In several cases, a 2mm difference in needle positioning resulted in a change of distal airway pressures of over 50%. Consistent reading of distal airway pressures from repeated runs indicates consistent needle positioning. A rigid tube was fitted to the trachea for each model so that th e needle could be aligned next to the tube to verify that the needle was correctly aligned to be concentric with the trachea. It was measured

PAGE 33

17 to be level with the surface of the desk, the printing platform, and the walls of the printer. The printer could repeatedly place the needle directly next to the alignment tube within a small fraction of a millimeter and parallel . We verified that the needle stayed parallel in directly next to the alignment tube as the needle moved towards and away from the t racheal opening. Signal and Data Processing T he system measures distal airway pressures within the airway block as a function of jet needle location in the trachea (dorsal ventral and left right) and depth into the trachea. For each position, the system is given a second of dwell time to allow residual vibration from moving to slow or stop before ventilation is engaged for three seconds. A MATLAB program was written to show live views of the data, make recordings, demultiplex the recorded signal s , plot th e data, filter noise, and analyze the resulting data. Signal analysis include d the mean pressures at each sensor for each depth, standard deviation, variance, homogeneity across sensors, and homogeneity across depths. A GUI was written to display data and conveniently export plots or excel data. We selected seventeen Honeywell transducers to record our pressures. We used HSCDANN 001PG AA5 1 psi high precision analog gauge pressure transducers and SCDANN001PDAA5 1 psi high precision analog differential pressure transducers to measure distal airway pressure . A Honeywell HSCDANN060PGAA5 60 psi high precision analog gauge pressure transducer was used to measure the pressure in the supply line. A 1.5 cm segment of silicone tube was affixed to each sensor to allow sensor to be easily attached and removed for calibration. The output signal from each passed through a Microchip Technology MCP6282 -

PAGE 34

18 E/P ND op amp to act as a buffer to reduce noise . The output from the op amp is routed through a n RC low pass filter set to 210 Hz . This frequency was set as an anti aliasing filter before the data was recorded by the DAQ. The sensors we re connected to a National Instruments USB 6009 eight channel , 14 bit data acquisition system ( DAQ ) which wa s then routed t o a computer .

PAGE 35

1 9 Figure 7 : Wiring diagram for the data acquisition system from the pressure transducers to the computer .

PAGE 36

20 A 4:1 multiplexer was used to increase the number of signals that could be acquired with the DAQ. A two bit binary counter was used to cycle the multiplexing chips. During each data acquisition sequence, 17 pressure signals and 2 binary counting signals were measured. The 17 pressure measurements include the 16 distal airway pressure sensors and the i nlet pressure. The binary count signals were continuously acquired on their own channels to allow sensor identif ication during demultiplexing . Each of the 16 airway pressure signals was acquired at 6.144 kHz for 1 ms out of every 4 ms (25.152 kHz asynchron ous sampling for all DAQ input channels) . The inlet pressure transducer had its own channel for uninterrupted data acquisition. The analog circuit used an Nexperia USA Inc. 74HC4052D,653 multiplexer chip and a 555 timer (960 Hz) connected to a Texas Inst ruments SN74LV393APWR four bit co unter , set as a two bit counter to control the switching . A 555 timer was set to run in a self triggering, bistable mode and was set to 960 H z . This 960 Hz binary count switched the MUX between four sensor inputs giving a final effective sampling rate of 240 sampling sets per distal airway sensor, per second or approximately 1500 samples per channel, per second . T he MUX cycled acquisition betwee n the 4 sensors on each channel and, f or each sensor, a sampling set was recorded that contained six to seven data points over the course of approximately one millisecond , followed by three milliseconds of no data for the sensor . During the demultiplexing process, the software use d the binary data to assign input data from each channel to the corresponding pressure signal. The data from the multiplexed signals would occasionally be recognized as part of the wrong channel because the a synchronous reading from the DAQ caused off timing between the binary count channels and the data

PAGE 37

21 readings. An off count data point would show up as a spike containing a single data point. Th e data was processed with a 10 th order median filter to remove t he spikes. A 9 th order IIR 200 Hz low pass filter was also applied to the pressure data. Airway Pressure Measurement Procedure The automated test apparatus ventilated the airway tree in thirteen different positions in the tracheal cross section. E ach posi tion in the trachea was considered a different test run and data were recorded at 7 different depths during each run. Figure 8 shows the thirteen different positions in the trachea l cross section that were tested. The block was ventilated in the center of the trachea and three steps in each direction. The outward st eps end ed with the center of the needle 2mm from the wall of the trachea. This provided a gap of just under 1 mm between the tracheal wall and the outside wall of the needle. The step spacing leading to the wall was set to 2 mm. For both the CT and idea lized airway block s we chose to use a uniform step distance and the same final distance from the wall rather than evenly spaced steps . The two tracheas had somewhat different shapes and diameters and we hypothesized that fluid dynamic surface effect s at the jet tracheal wall interface would affect homogeneity and would be a function of distance from the tracheal wall.

PAGE 38

22 Figure 8 : A top down view of the trachea with the thirteen ventilation positions shown: The center, and three steps in each direction. Flow Measurement We developed a venturi flow measurement device ( Figure 9 ) to allow the jet ventilation system to be set to a specific flow rate. Most jet ventilation systems start by setting the pressure in the supply line before the needle but needle gauge, length , and dynamic losses (e.g. friction in the tube , viscous losses from an elbow ) can substantively change the flow rate produced at the same pressure.

PAGE 39

23 Figure 9 : Venturi f low m eter d esign . The locations of A 1 , A 2 , P 1 , and P 2 (Eq. 1) are labeled in the cutaway (A) . Most ventilation flow meters currently on the market measure closed system flow which is relatively trivial. Jet ventilation is an open system, causing outside air to be entrained into the jet flow and pushed in to the airways. This means that the total flow may be many times higher than the flow coming out of the jet needle. One study noted a 20 fold increase in flow rate from a jet inserted in a laryngoscope [27] . This entrained air substantially changes the

PAGE 40

24 flow dynamics inside the trachea during ventilation and mus t be considered for an accurate reading. The amount of air entrained is based on several factors which are controlled by t he needle mounting system shown in Figure 10 . Figure 10 : Jet n eedle mounting system In order to record repeatable measurements, t he jet needle must be concentric and parallel to the venturi tube, stable enough to avoid vibrational or oscillatory effects in the jet, and fl ush with the entrance of the tube. Our system uses a mount to hold the needle concentric and parallel to the venturi flow meter and the mounting system is rigid, to prevent vibration or movement in the needle. 3D printing was used to manufacture the vent uri flow

PAGE 41

25 meter. To smooth out the texture in the printed plastic, an acetone vapor bath was constructed. The print was placed in an acetone vapor filled chamber for approximately 30 minutes and allowed to sit while the acetone softened the plastic. The a cetone source (paper towel damp with acetone) was removed and the chamber was opened and allowed to clear. The print was given approximately an hour to re harden before it was removed from the chamber. The venturi flow meter requires a hydrodynamically fully developed flow to function correctly . The jet produces a highly turbulent, undeveloped flow which requires either a long straight section or a flow conditioner to produce a fully developed flow before reaching the converging section of the tube . We used a long entrance length to develop the flow. The flow was calculated from the measured pressure difference at ports P 1 and P 2 ( Figure 9 ) using the Bernoulli equation (1) where flow (Q) is a function of the cross sectional area of the entrance (A 1 ), the cross sectional area of the constriction point (A 2 ) , the pressure before the entrance (p 1 ) , the pressure at the constriction (p 2 ) , and the air density ( ). The curve for the flow rate compared to the measured voltage is shown in Figure 11 .

PAGE 42

26 Figure 11 : The curve of the flow rate vs the measured voltage. This provides an accurate measurement from a flow rate of 0.25 L/s to approximately 3.25 L/s. Calibration of the meter was conducted with a calibrated 3 L syringe. T he flow meter was connected to the DAQ to record Q as 3 L of air was pushed through the meter at different flow rates . The integral of the flow rate during th e recording was the total flow through the meter and should be 3 L for each test. Additional Considerations Our mounting system is capable of holding multiple different needle gauges and lengths and accepts the standard needle mount bracket that will be used to subsequently mount the

PAGE 43

27 needle in the laryngoscope. We also designed a stable, twist lock mounting bracket shown in Figure 12 . This allows the bracket to be quickly and easily mounted in stable and repeatable fashion. Figure 12 : Twist on Mounting Bracket The needle mount is easily removed from the device and could be made as a plastic disposable piece or metal reusable part that can be sterilized in an autoclave. This is necessary because the mount will be in direct contact with the ventilation needle that will be inside the patient and therefore must be sterile. Our system includes a large, 16 pin, 5 volt, 16x2 backlit LCD display module that gives the real time flow rate as well as an average flow rate over the last several seconds of reading. We used a n Arduino to run the code and a custom built circuit to amplify and filter the sensor data from the pressure transducer. It displays a live view of the

PAGE 44

28 instantaneous flow rate and a running average of the flow for the three preceding seconds. Figure 13 shows our functioning prototype of the device without the housing and with a simplified mounting bracket. Figure 13 : Functioning prototype of the air flow entrainment measuring device, with housing removed

PAGE 45

29 CHAPTER IV R ESULTS We first consider jet flowmeter characterization and accuracy. The effect of n eedle position in the trachea i s explored and relationships between inlet pressure, flow rate, and distal airway pressure are defined . Finally, the effects of a laryngoscope and subglottiscope on the distal airway pressure are described . Flowmeter Characterization 15 flow rates were tested ranging from 0.24 L/s t o 2.3 L/s and the meter showed good accuracy across the range tested ( Figure 14 ) . At flow rates above 0.7 L/s the mean measured volume was 2.99 liter s with a standard deviation of 0.02 liters or 0.7% of the total volume. The mean error was 0.006 liters or 0.2%. The flowmeter was less accurate for flow rates below 0.7 L/s , reading an average volume of 3.14 liters with a standard deviation of 0.037 lit ers for a mean error of 4.7% with a standard deviation of 1.2%. Since the flowmeter was accurate at flow rates above 0.7 L/s but read consistently high at low flow rates a nonlinear calibration curve could be used to further improve accuracy. However, t h is was not necessary since our experiments , and clinical jet ventilation, us e flow rates above 1.2 L/s and this places the measurements in the high accuracy range of the meter.

PAGE 46

30 Figure 14 : V olume determined as the integral of the flow measured using the venturi flowmeter (blue circles) . The target value for each measurement is 3 L (red line) that is delivered using a calibrated syringe. Airway Pressure Measurements Each experimental run ventilated the airway block at one posit ion in the tracheal cross section and seven depths into the trachea . For each depth, data was simultaneously recorded at 16 different distal airways positions . Figure 15 shows the raw voltage reading from the 17 sensors with the jet needle 6 mm from the left wall of the trachea in the CT airway block . The first ventilation depth (D1) is 1 cm past the location of the vocal folds in the human subject, and each step was one centimeter deeper into the trachea. The final depth (D7) is 5.9 cm above the carina. In the idealized airway block, D7 is 5.5 cm above the carina. The low voltage (and low

PAGE 47

31 pressure) times are when the apparatus is repositioning the jet n eedle. The high voltage (and high pressure) times are when the jet ventilation is applied. Note that the voltage pressure relationship is different for the inlet pressure (black), channels 1 8 , and channels 9 16 . Figure 15 : The demultiplexed, unfiltered, uncalibrated voltage readings from each sensor during a ventilation test run on the CT airway block . The needle was centered and moved through all 7 depths . D1 D7 denote the depth into the trachea from 1cm to 7cm. S ensor specific calibration curve s were used to convert the raw voltage readings to the pressure readings shown in Figure 16 a . Each line represents the demultiplexed, median filtered pressure at one distal airway. These data are summarized in Figure 16 b where the mean distal airway pressure across all sensors is shown with a heavy line and the range from minimum to maximum pressure across all sensors is shown with a shaded area. This shows the homogeneity of ventilation at a gi ven depth and radial position in the tracheal cross section. Areas where the shaded range is small indicate homogeneous ventilation while p ositions with

PAGE 48

32 heterogeneous ventilation show wide spreads in the shaded area .

PAGE 49

33 Figure 16 : Distal airway pressure measurements in the CT airway block with the jet needle 6mm from the left wall of the trachea ( first step to the left ) showing the filtered pressure for all 16 distal airway sensors (a). The homogeneity graph (b) shows the mean pr essure (heavy black line) and the shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. Note that (b) is normalized by the inlet pressure . D1 D7 indicate depth into the trachea from 1cm to 7cm. D ata Normalization Distal airway pressures were divided by inlet pressures to define the normalized pressure, shown in equation 2, where Paw (n) is the pressure in airway (n) and P In is the inlet pressure. This sc aling was performed because the inlet press ure varied over time. Given these variations, i t would seem intuitive to find a more consistent regulator . However, we elected to keep the regulator because it was substantially similar to the actual regulator used in surgical ventilation. The pressure r egulator provide s an initial spike in pressure and then decay s to less than 9 0 percent of the peak value over a few seconds as shown in Figure 17 . The initial experimental configuration had additional pressure transients due to the uneven pressure p rovided by the compressor. That artifact was eliminated by the addition of a 10 gallon auxiliary air tank in series with the compressor. (2)

PAGE 50

34 Figure 17 : Transient in inlet pressure over the three second ventilation burst provided during the experiment . This is measured by a pressure transducer placed immediately before the ventilation needle. There were irregularities in the inlet pressure primarily from the regulator, however there was also a long timescale pressure decay that we attribute to changes in auxiliary tank pressure and short timescale transients that we attribute to inlet tubing compliance . These transient values in the pressure supplied to the ventilation needle were reflected in the flow rate and therefore the distal airway pressure , as they would in an actual jet ventilation setup. The data acquisition system had a con tinuous reading from the sensor recording the pressure at the ventilation needle. This reading was used as the reference for normalizing the data. By dividing the distal airway pressures by the inlet pressure, the effect of th ese transients

PAGE 51

35 were reduced to facilitate further analysis of the measurements Airway Pressure Relationships One of our primary interests was determining the distal airway pressure as a function of inlet pressure or flow, and several equations were deriv ed to relate the values. Since we have a consistent ventilation setup, any given inlet pressure correlates directly to a single flow rate and therefore we can translate inlet pressure to flow rate using the Darcy Weisbach equation. For more information on the flow measurement device, see the device design considerations in §3.3. We were consequently able to create equations relating both inlet pressure and flow to distal airway pressure. Figure 18 shows the relationship between total flow r ate and distal airway pressure. This is consistent across different depths, as shown in Figure 19 .

PAGE 52

36 Figure 18 : Relationship between distal airway pressure and total flow rate including entrained air. This is in the center position of the idealized airway block at 1 (a) , 4 (b) , and 7 cm (c) into the trachea.

PAGE 53

37 Figure 19 : Relationship between distal airway pressure and total flow rate shown at 1, 4, and 7 cm into the trachea, with the needle centered in the idealized airway block . Because clinical jet ventilation is guided by inlet pressure we have performed the remainder of the analysis using that parameter. Figure 20 shows the st rong linear correlation between inlet pressure and distal airway pressure for the idealized airway block .

PAGE 54

38 Figure 20 : Distal airway pressure vs inlet pressure for the center needle placement in the idealized airway model at depths of 1 (a), 4 (b), and 7 cm (c) . Figure 21 and Figure 22 show the inlet pressure v ersus d istal airway pressure for the idealized and CT airway mode ls , respectively . In both cases there is a strong linear correlation between inlet and distal airway pressure.

PAGE 55

39 Figure 21 : The linear regression of inlet pressure vs distal airway pressure for the CT airway model with the needle centered in the trachea at depth 1 . The CT airway model has a linear relationship between distal airway pressure (P distal ) and inlet pressure (P in ) ( Eq. 3 ) and R 2 = 0.9977 as shown in Figure 21 . The best fit to these data suggest s that a distal airway pressure of 0 is achieved when the i nlet pressure is 500 cmH 2 O . However, a t pressures below approximately 700 cmH 2 O there was a nonli near section that tapered off and a distal airway pressure of 0 was measured at an inlet pressure of 30 cmH 2 O . A figure is not included as this is well below the pressure used in ventilation . As such, Eq. 3 should only be used for inlet pressures above 700 cmH 2 O. (3 )

PAGE 56

40 Figure 22 : The linear regression of inlet vs distal pressure for the idealized airway model with the needle centered in the trachea at depth 1 . For the idealized airway model, there was a linear relationship between P distal and P in (Eq. 4 ) with R 2 = 0.9985 ( Figure 22 ) . Again, there was a nonlinear section for P in < 250 cmH 2 O and equation 4 should not be applied below this threshold . (4 ) Airway Pressure Distribution H omogeneity grids were created to describe the distribution of distal airway pressures. Figure 23 shows the numbered location s of the sensor s on the idealized airway block, shown

PAGE 57

41 from the inferior end of the lung block looking up . Figure 24 shows the corresponding homogeneity grid, which represents a view from the superior end of the block looking down the trachea. Figure 23 : The numbered positions o f the sensors on the idealized airway block

PAGE 58

42 1 5 9 13 2 6 10 14 3 7 11 15 4 8 12 16 Figure 24 : Sensor represented on the homogeneity grid. The bifurcations in the idealized airway tree rotate 90 degrees each generation. The first generation bifurcation is left right, the second generation bifurcation is anterior posterior , and so forth . For example, sensor 9 represents the airway that goes right at the first bifurcation, anterior at the se c ond bifurcation , left at the third bifurcation, and up at the fourth bifurcation, as shown in Figure 25 .

PAGE 59

43 9 Figure 25 : Example of airway bifurcations to reach sensor 9. Each darker shade represents a branch farther into the block. Figure 26 shows homogeneity grids for the needle centered in the trachea at depths one, four, and seven for the idealized airway model. With the needle centered, t he idealized airway model did not show a s ubstantial change in pressure with respect to depth of needle placement. At depth seven (5.5 cm from the carina), even a millimeter of movement towards one of the tracheal walls could result in heterogeneous ventilation (data not shown) .

PAGE 60

44 Figure 26 : Comparison of homogeneity in the distal airways for the idealized airway model at depths 1 ( a ) , 4 ( b ) , and 7 cm ( c ) . For the CT airway model, the trachea was curved, the diameter was variable , and it had an irregular cross section that changed with depth . This made it much more difficult to determine what the center of the trachea was, especially for a system designed to move in a straight line through the trachea. The airway homogeneity was sensitive enough to the position

PAGE 61

45 of the ventilation needle that a few millimeters of difference from the true center would make a clinically relevant impact in homogeneity. The normalized pressure distribution shown in Figure 27 was measured with the jet needle centered in the trachea l opening and downward in a direction parallel to the tracheal centerline . Figure 27 : The homogeneity grids at depths 1 ( a ) , 4 (b) , and 7 cm (c) with the jet needle centered in the trachea in the CT airway block .

PAGE 62

46 The CT airway block ( Figure 27 ) shows a marked decrease in distal airway pressure and increased pressure heterogeneity as d epth into the trachea increased. This is in contrast to the idealized airway model which showed onl y modest increased in heterogeneity and pressure with increased depth. With the needle situated closer to the tracheal wall ( Figure 28 ) , the pressure was more likely to favor a specific bifurcation, and the model typically showed bimodal ventilation where one region had higher pressures than the other . Trimodal ventilation wa s observed in the CT airway block ( Figure 28 c ) when the needle was positioned along the right wall of the trachea at a depth of 7 cm. The same position for the idealized airway model is available in the appendix for comparison.

PAGE 63

47 Figure 28 : Trimodal ventilation demonstrated in the CT airway block with the needle placed in the far right position in the trachea at depths of 1 (a), 4 (b), and 7 cm (c) . Figure 29 : The homogeneity grid at depth 1 with an 18 gauge needle in the center of the trachea for the CT airway tree.

PAGE 64

48 The diameter of the jet needle alters the flow rate and, as such, has a profound effect on distal airway pressure. Figure 29 shows the homogeneity grid for the CT airway block with the needle center ed in the trachea at a depth of 1 cm. However, in this case the 14 gauge needle was replaced with an 18 gauge needle. With the same inlet pressure setting, changing the needle from 14 ( Figure 27 ) to 18 gauge ( Figure 29 ) reduced the flow rate by 50.4% from 1.25 L/s to 0.62 L/s. This resulted in a n ormalized distal airway pressure decrease of 65.7% which is in accordance with the relationship shown in Figure 19 . In nearly every ventilation case in both airway models the ventilation pressure in the distal airways became more heterogeneous a s the ventilation needle was moved farther into the trachea and/ or closer to the tracheal wall . The s ensitivity to distance from the tracheal wall increased with depth into th e trachea. This is likely due to the turbulent, undeveloped flow produced by the jet ventilation needle. The trachea, acting as a tube entrance length , will produce a fully developed flow regime given enough length to do so , and a longer entrance length w ill more effectively develop the flow. To demonstrate these trends, homogeneity graphs were crea ted as outlined earlier in this chapter (e.g. Figure 16 ) . Figure 30 below shows the homogeneity of ventilation in the idealized airway model at varied distances from the left side of the trachea.

PAGE 65

49 Figure 30 : Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left side of the trachea is changed . The dark line in the center for each position shows the combined mean of all t he distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wall. The idealized airway model did not show a decrease in pressure with increased depth into the trachea for the center or the first step to the left . A decrease in pressure with depth was present but not continuous for left 2 and there was a small but continuous decrease in pressure for left 3. Every needle placement except for center showed a decrease in homogeneity with increased depth and this effe ct was more pronounced as the needle moved closer to the tracheal wall. At a depth of 1 cm past the laryngeal folds, the homogeneity was

PAGE 66

50 similar for all distances from the wall and there was a difference of approximately 10 percent in the mean pressure be tween the highest and lowest value. As depth increased , there was a substantial difference in pressure homogeneity and mean pressure with placements closer to the tracheal wall . Figure 31 : Comparison of needle placements in th e idealized airway block for the center and the placement s closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual senso r at that depth. Figure 31 sho ws the homogeneity graph comparing the center placement and the step closest to the tracheal wall in each direction. For al l movement directions , the pressure was

PAGE 67

5 1 relatively homogeneous for proximal needle positions and the normalized pressure were approximately equal . At more distal needle positions homogeneity decreased and the homogeneity of distal airway pressure became m ore sensitive to small changes in needle placement. Figure 32 : Homogeneity in the distal airways of the CT airway block as a function of distance from the left side of the trachea. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wall. In the CT airway bl ock pressure and homogeneity tended to decrease with increasing jet needle depth ( Figure 32 ) . This trend was seen in the left and right placement and , to a lesser

PAGE 68

52 exte nt , in the ventral and dorsal placement. Figure 33 : Comparison of needle placements in the CT airway block for the center and the placements closest to the tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. Figure 33 shows the CT airway block with the needle placements closest to the tracheal wall in each direction, as well as the center placement. The homogeneity consistently decreased with increasing depth into the trachea. The pattern of pressure loss with distal placement of the ventilation needle seen in Figure 32 is vaguely present but not nearly as defined.

PAGE 69

53 Laryngoscopes and Distal Airway Pressure Laryngoscopes and subglottiscopes are used during upper airway procedures to hold the airway open and all ow unobscured access for surgical tools . Since these s urgical scopes are common and can greatly increase the air entrainment, it is important to study the effects they have on distal airway pressure. An Osso f f Pilling 230 mm adult male subglottiscope (Mod el 52245) was tested at depths of 1 cm and 3 cm ( Figure 34 ) below the vocal folds in the CT airway block using a standard Pilling C8 14 gauge ventilation needle. The end of the subglottiscope is typically inserted one centimeter past the laryngeal folds and the folds press against the outer walls to create a seal. Our experiment used putty to seal the subglottiscope to the opening of the airway block. The normalized pres sure averaged just slightly higher than without the subglottiscope and there was no physiologically relevant difference between scope placement depths.

PAGE 70

54 Figure 34 : Pressure homogeneity grids for the subglottiscope with the orifice 1 and 3 centimeters past the laryngeal folds. This is in the CT airway block. The ventilation homogeneity was also measured for an Ossoff Pilling adult male 168 mm micro laryngoscope (model 522191) with the standard Pilling C8 14 gauge needle. The laryngoscope was placed at a depth of 1 cm past the laryngeal folds and the needle was mounted on the bottom of the right wall with the end 2.5 cm into the laryngoscope as is common in clinical practice . Figure 35 shows the same laryngoscope, placed at the same depth, with the same needle mounting position, with the ventilation needle angled up within the laryngoscope at 10 , 5 , 3, and 0 degrees .

PAGE 71

55 Figure 35 : The homogeneity grids for the Ossoff Pilling Laryngoscope ventilation. The needle is mounted at the bottom of the right wall and tilted up at angles of 10, 5, 3, and 0 degrees . This is the CT airway model . In Figure 33 we saw a tre nd in the CT airway model of increased pressure as the needle moved superior in the trachea . This trend of superior needle placement and increase d flow rate resulting in higher distal airway pressures continued with the addition of the subglottiscope and laryngoscope . The laryngoscope moved the needle placement even farther

PAGE 72

56 from the lungs , a nd the mean pressure at every angle was higher than it wo uld be without the laryngoscope. By moving the ventilation needle from a 10 degree angle to a 3 degree angle, there was an 82% increase in distal airway pressure. The experimental needle placement was set by a surgeon in the same manner that it would be placed during an actual ventilation, and the 10 degree change in angles is within what would be typical movement during surgery.

PAGE 73

57 CHAPTER V D ISCUSSION The primary goal during mechanical ventilation is to exchange gas to maintain homeostasis without injuring the lung . There are three important pressures in ventilation : the peak inspiratory pressure (P IP ), plateau pressure (P Plat ) , and the positive end expiratory pressure (PEEP) . The peak inspiratory pressure (P IP ) is the highest pressure in the lungs during inhalation and is a combination of lung resistance and elastance. T he pressure at the end of inspiration is P plat and this is predominantly governed by the lung elastic recoil and lung volume . The PEEP is the pressure at the end of exhalation that is used to prevent airway and alveolar collapse. LFJV does not provide a PEEP. The tidal volume delivered to the lungs (Vt) = Cst (P Plat PEEP) , where Cst is the pulmonary system compliance . Gas delivery is defined by the minute ventilation which is the product of tidal volume and respiratory rate and is the amount of gas received per minute. The inspiratory and expiratory pressure that drive ventilation must be carefully selected to deliver sufficient gas w hile keeping P Plat low enough to avoid volutrauma and PEEP high enough to prevent damage via atelectrauma [20 23] . If all regions of the lung are provided with an appropriate range of inspiratory and expiratory pressure s, then the alveoli will be properly supplied with fresh gas to prevent hypoxia and hypercapnia . However, h igh P Plat can drive excessive v olumes of air into the lungs and, in the most severe cases, volutrauma will occur . It is also possible that regions of the lung are subjected to different ranges of inspiratory and expiratory pressures ( heterogeneous ventilation ) . I f the spatial mean distal airway pressure across the lung is at a reasonable level

PAGE 74

58 but one airway is receiving excessive pressure , this will r esult in excessive volume an d an increased risk of volu trauma in the parenchyma supplied by that airway . If the spatial mean di stal airway pressure is reduced to avoid overdistension in the high pressure region, then the remainder of the lung may receive insufficient flow resulting in impaired gas exchange . By providing homogeneous ventilation , the spatial maximum of the pressure can be reduced while maintaining a higher mean distal airway pressure, allowing for appropriate ventilation to prevent hypoxia and hypercapnia with a reduced risk of complications in the lungs. The rate of complications specific to LFJV are not well established and consequently, we do not know the optimal settings. Data and statistics from other types of jet ventilation can be used as a general reference for what we might expect to see in LFJV, but should not be considered the same as actual dat a on LFJV. In high frequency jet ventilation, complications involving gas exchange, including hypoxemia and hypercapnia, occur in 18 and 22% of adult bronchoscopy jet ventilation cases respectively and serious complications including barotrauma, cervical e mphysema, and tension pneumothorax occur in approximately 1% of adult bronchoscopy cases [49 51] transtracheal jet is usually hastily set up and complications are very common, with a barotrauma rate of 32% o f cases and a total complication rate, including hypoxia and hypercapnia, of 51% [31] . Accordingly, we have used idealized and patient specific phys ical models to investigate how to best prevent hypoxia, hypercapnia, and lung trauma. The Relationship Between Inlet Pressure , Flow Rate, and Mean Distal Airway Pressure Clinical care providers currently use pressure at the ventilation needle as a ref erence for

PAGE 75

59 jet ventilation. Establishing how pressure at the needle relates to distal airway pressure may help providers avoid VILI . The most fundamental parameter for invasive mechanical ventilation is the tidal volume which, as discussed above, is a fu nction of P Plat . There is no clear correlate in clinical jet ventilation settings so we have established the relationship between inlet pressure (P in ) and distal airway pressure (P Distal ) . We found a linear relationship between inlet pressure and distal airway pressure ( Figure 21 and Figure 22 ) . We established Equation 3 and Equation 4 for the ventilation models and found that the idealized airway model and CT airway model had nearly the same slope but different intercepts. These intercepts were extrapolated from th e linear best fit s that were based on P in > 700 cmH 2 O for the CT airway model and P in > 250 cmH 2 O for the idealized airway model . T he relationship between inlet and distal airway pressures has a nonlinear section at very low inlet pressures. However, this nonlinearity is well below the pressures seen in clinical use . In the low pressure, nonlinear regime almost no ne of the pressure makes it to the distal airways until a critical pressure is reached. At that critical pressure, the line ar section starts and a consistent portion of the inlet pressure makes it to the distal airways. A irway geometry affects the inlet pressure needed to leave the nonlinear section and start providing significant pressures in the distal airways. For the CT airway model the nonlinear section goes from 0 to 8 psi and in this section very little pressure makes it to the fifth generation airway. For the idealized airway model, the nonlinear section goes from 0 to 3 psi. The slope of the linear section was cons istent between the blocks despite the nu merous difference in geometry. However, the re is no inherent relationship between inlet pressure and P distal ; this

PAGE 76

60 correlation is contingent upon the physical configuration of the mechanical ventilation apparatus . For example, c hanges in needle gauge will affect flow rate and the flow rate is directly reflected in P distal . Figure 29 shows the pressure in the distal airways with the same parameters as Figure 27 a but the 14 gauge needle was switched for an 18 gauge needle. This resulted in a 50.4% decrease in flow rate. The same inlet pressure and position showed a 65.7% decrease in distal airway pressure because of the reduction in flow rate. We assessed the system at three different depths, one, four, and seven centimeters into the trachea and found a 2 nd degree polynomial relationship between the flow rate and distal pressure. The equations describing the relationship are seen in Figure 18 . Inlet press ure can only be used as a reference for ventilation when it can be reliably related to flow rate. This means that major and minor fluid dynamic losses must be controlled for. These sources include needle parameters such as gauge, length, and any bends or curves in the needle; supply hose parameters such as diameter, coupling type, compliance, and length past the pressure regulator; and other sources of loss in the system including valve size, shape, and locations. These sources of loss must all be kept c onstant between tests so that the same pressure will provide the same flow rate. During a surgical procedure, different providers have different preferences in laryngoscopes, needle types, needle mounting positions, and overall setup. It is highly unlike ly that the variables responsible for dynamic losses would be consistent from procedure to procedure. Based on these findings, we assert that the pressure of the supply line should not be used as the reference for jet ventilation.

PAGE 77

61 Jet Flowmeter Developme nt and Characterization In order to address the fact that P in and P Distal are not correlated across different ventilation hardware configuration s we have developed a flowmeter to help clinicians determine appropriate settings for their patients. Inline f low meters currently on the market can accurately determine the flow rate coming out of the ventilation needle . However, since LFJV is an open system t he total flow rate delivered during jet ventilation must include both the flow coming out of the needle and additional outside air that is entrained into the flow. In some cases, the entrained flow is a small portion of the total flow. In other cases, the entrained flow makes up the vast majority of the total flow [27] . The ratio of flow from the needle to entrained flow is dependent on needle gauge, jet exit velocity, and other jet flow properties such as turbulence, and interactions with other surfaces, such as a laryngoscope. Our device allows for accurate readings at a range of flow rates and includes air entrainment. We created a characterization curve ( Figure 14 ) for the meter and found that it has an error of less than 6% across the full range of flow rates . A bove roughly 0.75 L /s the meter had an error that rarely exceeded 1%. For our experiment, which typically used 1.2 L/s to 1.4 L/s , we had a highly accurate device. The curve showed a non linear trend in the data points and the data was tightly clustered along this curve. By creating an adjustment curve for calibrating the meter, we estimate that the error could be reduced to less than 1% across the full range of flow rates. This device would not be able to replace the experience of the anesthesiologists in determining the proper flow setting since this parameter varies with patient physiology. For

PAGE 78

62 example, a pediatric ventilation case would need significantly less flow than an adult case. Many pathologic conditions would require adjustment of flow rates and e ven the use of different laryngoscope s would change the required flow settings. Instead, this device is meant to be a tool to grant anesthesiologists the ability to provide repeatable flow settings. With the current standard of care, even if a clinician knew that on e laryngoscope provided higher distal airway pressures, trying to accurately reduce the flow by feel alone is not feasible. The ability to produce consistent flow rates from the needle and knowledge specific to the scope being used during a procedure coul d be extremely valuable to an experienced anesthesiologist. Jet Needle Position and Distal Airway Pressure Heterogeneity The radial and axial position of the jet needle in the trachea are key determinants of distal airway pressure. The relationships betwe en depth, distance from the wall, and distal airway pressure is of major importance in emergency jet ventilation where a laryngoscope or subglottiscope is not typically used. For the sake of real time measurement and clinical relevance, rather than normal izing to flow rate, we normalized to inlet pressure. Pressure is currently the clinical norm for setting up jet ventilati on system along with feeling the jet from the needle . By using the same needle and ventilation setup every time, we knew that flow ra te was a simple function of pressure. I n the CT airway model , a s depth into the trachea increased there was generally a counterintuitive decrease in distal airway pressure and a decrease in ventilation homogeneity . In the idealized airway model, there were no physiologically relevant changes in mean distal airway pressure with increased jet needle depth . However, the degree of heterogeneity

PAGE 79

63 increased with jet needle depth . We looked at homogeneity between the different airways in the lung at during the same ventilation burst and how changing the position and pressure of the needle would affect the pressure at a given location in the lung . We found that surface effects along the wall of the trachea caused a negative correlation between the distance from the tracheal wall and homogeneity of ventilation. At shallow depths into the trachea, a fully developed flow regime was able to form regardless of distance from the wall. As a result, a deeper needle placement made the model more sensitive to slight variations in needle position. A centered needle in the proximal trachea provided homogeneous ventilation more easily than a needle with a distal placement. This is clinically relevant for determining needle placement for optimal hom ogeneity w hich in turn will improve gas exchange a nd reduce the risk of volu trauma and other forms of ventilator induced lung injury. In a clinical setting, the ventilation needle should ideally be centered in the trachea and as far from the carina as possible. Due the physical and visual obstruction that would be caused by having the needle centered, this is not possible during surgery. It must be placed against one of the walls. Looking at Figure 31 and Figure 33 , we can see that at depth 1 the placement of the needle against the different walls makes no m ore than a 15% difference in the spatial mean airway pressure while homogeneity is consistent between positions. Effects of a Laryngoscope and S ubglottiscope on Distal Airway Pressure Scopes are typically used during LFJV. They can significantly change the flow rate and jet dynamics resulting in altered distal airway pressures. A trend of increasing pressure with

PAGE 80

64 proximal placement of the needle seen in Figure 34 and Figure 35 continued with the addition of the subglottiscope and laryngoscope. The extended entrance length from the laryngoscope contributed to homogenous ventilation by allowing a more fully developed flow regime . Also in Figure 34 and Figure 35 w e saw a 20% increase in distal airway pressure with the subglottiscope and at least a 65% increase in distal airway pressure with the Ossoff Pilling Laryngoscope as compared to the centered needle at depth one, shown in Figure 27 . Two depths, one centimeter and three centimeters, past the laryngeal folds were tested for the subglottiscope and there was no difference in distal airway pressure or homogeneity. This mea ns that accidentally pushing the subglottiscope a few centimeters farther than normal past the laryngeal folds during surgery would not make a substantial difference in ventilation. A surgeon familiar with the ventilation technique set up the Ossoff Pillin g laryngoscope in a typical fashion and there was just over 10 degrees of movement in the needle within the laryngoscope . The needle was mounted at the bottom of the right side of the rectangular opening to the laryngoscope. The screw mount used to hold the needle in place prevents the needle from sliding, but allows for pivoting easily. The weight of the ventilation hose was enough to pull down on the end of the needle until the proximal end of the needle hit the laryngoscope, resulting in an upward til t of 10 degrees. By lifting the supply hose, the needle effortlessly tilted until the tip of the needle hit the laryngoscope resulting in a level needle placement. The weight of the supply hose was the primary resistance to needle movement. Those 10 degrees made a remarkable difference in distal airway pressure. There was an 82% increase in distal airway pressure by moving the needle across a 7 degree arc from 10

PAGE 81

65 degrees to 3 degrees . An 82% increase in pressure could be the difference betwe en under ventilation and barotrauma. A n inadvertent 7 degree needle movement could happen during a surgical procedure when tools are placed in the laryngoscope, when the evacuator is adjusted, or if the surgeon is simply unaware of the difference in press ure and bumps the needle. This means that for many jet ventilation cases the distal airway pressure will occasionally double at random intervals throughout the surgery unbeknownst to the surgeon or anesthesiologist. Unlike the laryngoscope, the subglotti scope had a secure mount for the ventilation needle and it was not possible to adjust the needle position. This promotes consistent distal airway pressure. Conclusions and Suggestions for Clinical Practice The current standard for low frequency jet ventilation has done many things well, but could be greatly improved with a few small changes. A quantitative measurement of total entrained flow would help anesthesiologists provide consistent, repeatable flow rates. The correlati on between needle flow rate and total delivered flow rate for each laryngoscope should be established, and an equation or table should be provided with scopes to help anesthesiologists provide an ideal distal airway pressure. Future experiments should loo k at trends in different laryngoscopes and the appropriate flow rate for each one. Stable n eedle mounts should be built into the laryngoscope so the needle cannot be easily bumped, resulting in pressure spikes and drops. Improvements to laryngoscope desi gn and flow measurement could provide more easily controlled, consistent distal airway pressures, to improve gas exchange and reduce the risk of trauma to the lungs.

PAGE 82

66 Future Investigations The models designed for this experiment could be used for many follow up experiments. Characterization of additional laryngoscopes and subglottiscopes could show how consistent distal airways are within the same type of scope. Perhaps additional types of endoscopic tools such as bronchoscopes could be tested. It wo uld also be interesting to run the same tests on additional models. Having several healthy adult models could lend additional credibility to the study and show the variety or consistency between people. Pathologic or pediatric models would also be worth investigating. The differences in equations, regression lines and data could be insightful. It might be worth adding balloons or a compliant outer section to the model to see how that effects the fluid dynamics of the block. A particle image velocimetry experiment could elucidate what is actually happening in the airways during ventilation, especially at the carina. Perhaps this would explain why there was a counterintuitive decrease in the distal airway pressure on the same side as the needle placement. An animal model may ultimately be the most realistic form of this experiment. Placing transducers into the distal airways and measuring the actual pressures during ventilation could be insightful.

PAGE 83

67 REFERENCES 1. Tucker, W.D. and B. Burns, Anatomy, Thorax, Heart Pulmonary Arteries , in StatPearls . 2019. 2. Donley, E.R. and J.W. Loyd, Anatomy, Thorax, Wall Movements . 2019. 3. Patwa, A. and A. Shah, Anatomy and physiology of respiratory system relevant to anaesthesia. Indian journal of anaesthesia, 2015. 59 (9): p. 533 541. 4. Verschakelen, J.A. and W. De Wever, Basic Anatomy and CT of the Normal Lung , in Computed Tomography of the Lung: A Pattern Approach , J.A. Verschakelen and W. De Wever, Editors. 2018. p. 3 19. 5. G overnment, T.U.S., Lung Diagram. 2006. 6. Macklem, P.T., The Mechanics of Breathing. American Journal of Respiratory and Critical Care Medicine, 1998. 157 (4): p. S88 S94. 7. Saraswat, V., Effects of anaesthesia techniques and drugs on pulmonary function. I ndian journal of anaesthesia, 2015. 59 (9): p. 557 564. 8. Knowlson, G.T.G. and H.F.M. Bassett, THE PRESSURES EXERTED ON THE TRACHEA BY ENDOTRACHEAL INFLATABLE CUFFS. British Journal of Anaesthesia, 1970. 42 (10): p. 834 837. 9. Smallwood, C.D. and B.K. Wals h, Noninvasive Monitoring of Oxygen and Ventilation. Respiratory Care, 2017. 62 (6): p. 751. 10. Broche, L., et al., Dynamic Mechanical Interactions Between Neighboring Airspaces Determine Cyclic Opening and Closure in Injured Lung. Critical Care Medicine, 2017. 45 (4): p. 687 694. 11. Motta Ribeiro, G., et al., Spatial Heterogeneity of Lung Strain and Aeration and Regional Inflammation During Early Lung Injury Assessed with PET/CT. Academic Radiology, 2019. 26 (3): p. 313 325.

PAGE 84

68 12. Michiels, C., Physiological and pathological responses to hypoxia. The American journal of pathology, 2004. 164 (6): p. 1875 1882. 13. S, P. and M. SH, Physiology, Carbon Dioxide Retention. StatPearls [Internet], (2019 Jan ). 14. Heinz, U.E. and J.D. Rollnik, Outcome and prognosis of hypoxic brain damage patients undergoing neurological early rehabilitation. BMC research notes, 2015. 8 : p. 243 243. 15. Gattinoni, L., et al., Relationships between lung computed tomographic density, gas exchange, and PEEP in acute respiratory failure. An esthesiology, 1988. 69 (6): p. 824 32. 16. Bilek, A.M., K.C. Dee, and D.P. Gaver III, Mechanisms of surface tension induced epithelial cell damage in a model of pulmonary airway reopening. J. Appl. Physiol., 2003. 94 : p. 770 783. 17. Kay, S.S., et al., Pres sure gradient, not exposure duration, determines the extent of epithelial cell damage in a model of pulmonary airway reopening. J. Appl. Physiol., 2004. 97 : p. 269 276. 18. Mead, J., T. Takishima, and D. Leith, Stress distribution in lungs: a model of pulm onary elasticity. Journal of applied physiology, 1970. 28 (5): p. 596 608. 19. Makiyama, A.M., et al., Stress concentration around an atelectatic region: A finite element model. Respir Physiol Neurobiol, 2014. 201 : p. 101 10. 20. Tremblay, L.N. and A.S. Slu tsky, Ventilator induced lung injury: from the bench to the bedside. Intensive Care Med, 2006. 32 (1): p. 24 33. 21. Hamlington, K.L., et al., Predicting ventilator induced lung injury using a lung injury cost function. J Appl Physiol (1985), 2016. 121 (1): p. 106 14. 22. Henderson, W.R., et al., Fifty Years of Research in ARDS.Respiratory Mechanics in Acute Respiratory Distress Syndrome. American

PAGE 85

69 Journal of Respiratory and Critical Care Medicine, 2017. 196 (7): p. 822 833. 23. Baumgardner, J.E., et al., Effec ts of Respiratory Rate, Plateau Pressure, and Positive End Expiratory Pressure on PaO2 Oscillations after Saline Lavage. American Journal of Respiratory and Critical Care Medicine, 2002. 166 (12): p. 1556 1562. 24. Smith, B.J. and J.H.T. Bates, Assessing the Progression of Ventilator Induced Lung Injury in Mice. IEEE Trans. Biomed. Eng., 2013. 60 (12): p. 3449 3457. 25. Otto, C.M., et al., Spatial and temporal heterogeneity of ventilator associated lung injury after surfactant depletion. Journal of applied physiology (Bethesda, Md. : 1985), 2008. 104 (5): p. 1485 1494. 26. Turner, J.S., Turbulent entrainment: the development of the entrainment assumption, and its application to geophysical flows. Journal of Fluid Mechanics, 1986. 173 : p. 431 471. 27. Van Der Spek, A.F., P.M. Spargo, and M.L. Norton, The physics of lasers and implications for their use during airway surgery. Br J Anaesth, 1988. 60 (6): p. 709 29. 28. Evans, E., P. Biro, and N. Bedforth, Jet ventilation. BJA Education, 2007. 7 (1): p. 2 5. 29. Bar aka, A.S., et al., Low frequency jet ventilation for stent insertion in a patient with tracheal stenosis. Canadian Journal of Anesthesia, 2001. 48 (7): p. 701 704. 30. Pathak, V., et al., Ventilation and Anesthetic Approaches for Rigid Bronchoscopy. Annals of the American Thoracic Society, 2014. 11 (4): p. 628 634. 31. Duggan, L.V., et al., Transtracheal jet ventilation in the 'can't intubate can't oxygenate' emergency: a systematic review. Br J Anaesth, 2016. 117 Suppl 1 : p. i28 i38. 32. Care, U.o.I.H. Jet V entilation Anesthesia Transoral for Laryngeal Surgery . Iowa Head and Neck Protocols 2019 6/18/2019];

PAGE 86

70 Available from: https://medicine.uiowa.e du/iowaprotocols/jet ventilation anesthesia transoral laryngeal surgery . 33. UCSF, D.o.A.P.C. Manual Jet Ventilation . 2013. 34. Reed, J.P., et al., Studies with transtracheal artificial respiration. Anesthesiology, 1954. 15 (1): p. 28 41. 35. Chhangani, S.V ., CHAPTER 30 Independent Lung Ventilation and Bronchopleural Fistula , in Mechanical Ventilation , P.J. Papadakos, B. Lachmann, and L. Visser Isles, Editors. 2008. p. 337 354. 36. Putz, L., A. Mayné, and A. S. Dincq, Jet ventilation during rigid bronchosc opy in adults : a focused review." . Vol. 2016. 2016. 37. Klain, M. and R.B. Smith, High frequency percutaneous transtracheal jet ventilation. Crit Care Med, 1977. 5 (6): p. 280 7. 38. Pawlowski, J., Anesthetic considerations for interventional pulmonary pro cedures. Curr Opin Anaesthesiol, 2013. 26 (1): p. 6 12. 39. Bouroche, G., et al., Rescue transtracheal jet ventilation during difficult intubation in patients with upper airway cancer. Anaesth Crit Care Pain Med, 2018. 37 (6): p. 539 544. 40. Noonan, M., et al., Intraoperative High Frequency Jet Ventilation Is Equivalent to Conventional Ventilation During Patent Ductus Arteriosus Ligation. World J Pediatr Congenit Heart Surg, 2017. 8 (5): p. 570 574. 41. Abad, H.L., M. Ajalloueyan, and A.R. Jala li, Impact of body mass index (BMI) on ventilation during low frequency jet ventilation. Otolaryngol Head Neck Surg, 2007. 136 (3): p. 477 80. 42. Altun, D., et al., Surgical Excision of Postintubation Granuloma Under Jet Ventilation. Turkish journal of ana esthesiology and reanimation, 2014. 42 (4): p. 220 222. 43. Philips, R., B. deSilva, and L. Matrka, Jet ventilation in obese patients undergoing airway surgery for subglottic and tracheal stenosis. Laryngoscope, 2018. 128 (8): p. 1887 1892.

PAGE 87

71 44. Goudra, B.G., et al., Anesthesia for Advanced Bronchoscopic Procedures: State of the Art Review. Lung, 2015. 193 (4): p. 453 465. 45. Bourgain, J.L., et al., [Guide for the use of jet ventilation during ENT and oral surgery]. Ann Fr Anesth Reanim, 2010. 29 (10): p. 720 7. 46. Florens, M., B. Sapoval, and M. Filoche, An anatomical and functional model of the human tracheobronchial tree. J Appl Physiol (1985), 2011. 110 (3): p. 756 63. 47. Horsfield, K., et al., Models of the human bronchial tree. J Appl Physiol, 1971. 31 (2): p. 207 17. 48. Weibel, E.R., Morphometry of the Human Lung , ed. E.R. Weibel. 1963, 1 151. 49. Fernandez Bustamante, A., et al., High frequency jet ventilation in interventional bronchoscopy: factors with predictive value on high frequency jet ventil ation complications. J Clin Anesth, 2006. 18 (5): p. 349 56. 50. Hautmann, H., et al., High frequency jet ventilation in interventional fiberoptic bronchoscopy. Anesth Analg, 2000. 90 (6): p. 1436 40. 51. Conacher, I.D., Anaesthesia and tracheobronchial sten ting for central airway obstruction in adults. Br J Anaesth, 2003. 90 (3): p. 367 74.

PAGE 88

72 APPENDIX Figure 36 : Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the dorsal wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mea n value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Dorsal 1 is closest to the center and dorsal 3 is closest to the tracheal wall.

PAGE 89

73 Figure 37 : Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the ventral wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors f or the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Ventral 1 is closest to the center and ventral 3 is closest to the tracheal wall.

PAGE 90

74 Figure 38 : Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is closest to the center and left 3 is closest to the tracheal wall.

PAGE 91

75 Figure 39 : Homogeneity in distal airway pressure in the CT airway block as a function of depth as the distance from the right wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. T he shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Right 1 is closest to the center and right 3 is closest to the tr acheal wall.

PAGE 92

76 Figure 40 : Homogeneity in distal airway pressure in the CT airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.

PAGE 93

77 Homogeneity in distal airway pressure in the idealized airway block as a function of depth as the distance from the left wall of the trachea is changed. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth. The steps toward the wall are numbered from the center of the trachea out. Left 1 is c losest to the center and Left 3 is closest to the tracheal wall.

PAGE 94

78 Figure 41 : Homogeneity in distal airway pressure in the idealized airway block as a function of depth in the position closest to tracheal wall in each direction. The dark line in the center for each position shows the combined mean of all the distal airway sensors for the three seconds of ventilation at that depth. The shaded area represents the range from the highest to lowest mean value for an individual sensor at that depth.

PAGE 95

79 Figure 42 : Homogeneity grid for v entilation in the idealized airway block with the needle placed in the far right position in the trachea at depth 1.

PAGE 96

80 Figure 43 : Homogeneity grid for v entilation in the idealized airway block with the needle placed in the far right position in the trachea at depth 4. Bimodal ventilation is seen here.

PAGE 97

81 Figure 44 : Ventilation in the idealized airway block with the needle place d in the far right position in the trachea at depth 7. Bimodal ventilation is seen here.

PAGE 98

82 DECLARATION OF ORIGINAL WORK Thesis is my original work. Further, I confirm that all writing is my own writing. Work from others has been cited appropriately. __ Joshua Pertile ___ ____ 7/19 /19 _____ Student Name Date